スカイロジック List of Products and Services

Even manufacturers that globally deploy valve parts are considering our inspection software. Our software is utilized in specialized and intricate products across various industries. 【Inspection Settings and Results】 The task of enclosing the parts to be detected in a rectangle for parameter generation is called annotation. Each part is registered with the Label name shown in the right image, and it is set up so that if there is even one defect, it will be deemed a failure. We would like you to consider stable production through the automation of inspections, especially for products that require a high level of solution. 【Software and Equipment Used】 Software Used: DeepSky Learning Edition Field of View: Approximately 78 x 62 mm Minimum Size of Inspection Target: 2 mm Number of Inspection Points: 1 point, detecting defects from the entire screen Camera Resolution: 1.3 million pixels Lens Focal Length: 12 mm Distance Between Lens and Product: Approximately 150 mm Lighting: Ring lighting Distance Between Lighting and Inspection Item: Approximately 100 mm above

• Image Processing Software

It seems that there are many washers and similar screws used in various industries, making it often cumbersome to distinguish between them. This time, we received an inquiry about how to differentiate washers and screws. Since this is a common issue, we have also created a video to make it easier to visualize. Please take a look for reference. "Discovering the simultaneous omission of two types of parts and stopping the conveyor" https://skylogiq.co.jp/DIY_HowTo/368 【Inspection Settings and Results】 We enclosed the target areas we wanted to detect with a frame and labeled them by type. In this verification, we focused on six types of parts to examine the feasibility of differentiation. We believe that false detections and non-detections occurred due to insufficient training data for the patterns in which the target objects appeared in the images, and we expect that simply increasing the patterns in the training data will improve detection accuracy. However, we cannot determine the final rate of false detections or non-detections until we actually conduct the training.

• Image Processing Software

The industrial knife manufacturer focuses on high quality, precision, and functionality in the production of cutting tools for various objects, and is committed to pursuing product quality control. This time, we are verifying the judgment of defects in cutting machine parts. The current inspection method is conducted using visual checks and jigs, but there tends to be negligence during busy periods, leading to incidents that result in complaints. Since this is a common case with a variety of workpieces, we have also created a video to make it easier to visualize. Please take a look for reference. "Discovering the omission of two types of parts at the same time and stopping the conveyor" https://skylogiq.co.jp/DIY_HowTo/368 【Inspection Settings and Results】 It was possible to detect defective parts. However, due to the variety of workpieces, it is necessary to set an appropriate field of view for each size. There were also instances where defects could not be captured directly from above, necessitating angled imaging of holes, which is expected to make positioning and environmental conditions more stringent, as reported. 【Software Used】 Software used: DeepSky

• Image Processing Software

This is a request for a free evaluation to distinguish discoloration in parts (washers). We received an inquiry from a manufacturer that produces everything from fine electronic components to large fire-resistant glass. While looking forward to future developments, we conducted a simple verification test to identify the washers as an entry point. In the free sample evaluation, we will first conduct a simple verification using the samples or images provided and report the results. During the simple verification, we will evaluate whether the desired detection/judgment can be achieved using our internal equipment. If detection or judgment is possible in the simple verification, we recommend conducting a test (feasibility verification) assuming actual operation, where we will evaluate processing time and judgment accuracy. If feasibility verification is to be conducted, you can choose to continue with our company (for a fee) or use our loaned equipment for verification at your company. [Inspection Settings and Results] We labeled the normal washers as OK and the washers with part of them painted black as NG. We annotated six images with changed positions for the time being. The detection was accurate. It is possible to set how many OKs are needed to pass or if there is any NG to fail. Other detailed settings allow for recognition point counts and area judgments.

• Image Processing Software

We receive numerous inquiries about counting parts. Recently, we developed software for conveyor counting as an advanced application of DeepSky. We proposed a method where, while inspecting and counting, if you want to count 1,000 items, you can set the conveyor to light up at 990, have the workers line up in a row, and then stop the conveyor again at 1,000. In this inquiry, we attempted a test to count 23 correct workpieces while stationary on the workbench. 【Inspection Settings and Results】 In the settings shown in the left image, the parts to be counted are labeled as "Work," while other mixed parts are labeled as "Excluded." When set up as shown in the image above, the configuration passes if there are 20 counted parts (Work). The "Rs" on the right indicates the number of recognition points to be detected, set individually. The right image displays the detection screen. 【Software and Equipment Used】 Software Used: DeepSky Field of View: 280 x 200 mm Minimum Size of Inspection Target: 2 mm Number of Inspection Points: 8 Camera Resolution: 1.3 Megapixels Lens Focal Length: 6 mm Distance Between Lens and Product: Approximately 330 mm Lighting: Indoor Fluorescent Light

• Image Processing Software

Metal products such as die-castings often reflect light, making inspection difficult in the past. With our inspection software DeepSky, released in 2020, it is now possible to operate even reflective metal products like die-castings with simple settings. 【Inspection Settings and Results】 By using DeepSky's inspection features, we were able to determine dents, impressions, and chip blockages in just 0.35 seconds. The image on the left shows the hole that should be present in a good product, while the image on the right is taken with the same color (white) as the good product's "hole," but it is accurately recognized as defective. The number at the top of the frame indicates the AI's confidence level (number of recognition points). 【Software and Equipment Used】 Software Used: DeepSky Learning Version Field of View: Approximately 56 x 42 mm Minimum Size of Inspection Target: 20 mm Number of Inspection Points: 1 overall Camera Resolution: 1.3 million pixels Lens Focal Length: 12 mm Distance Between Lens and Product: Approximately 150 mm Lighting: Flashlight (Chip blockage: spotlighting the processing hole)

• Image Processing Software

Our image inspection software is being considered in various fields and industries. The combination of AI (deep learning) and image inspection can be beneficial even in high-level solution areas. We recommend efficient inspections with DeepSky. [Inspection Settings and Results] The results of the verification conducted with the samples you provided showed that it was possible to detect "dents, impressions, chips, and large holes" and determine pass or fail. This time, we used a software called DeepSky, which utilizes AI (Deep Learning) for inspection. By training the software on the areas we want to detect, it can adjust its own setting parameters and recognize them. In actual operations, various patterns (sizes and features) of defects are expected, so it is necessary to increase the number of training images.

• Image Processing Software

We received samples from an aluminum processing manufacturer that we guided at the Sky Logic booth during the exhibition and conducted a simple evaluation. We attempted to detect "defects" and "scratches" on the workpiece. 【Inspection Settings and Results】 However, inspecting only one side does not allow for the inspection of all surfaces, so we proposed a method of moving the inspection item while changing the position of the lighting. We mounted a fixture on a rotary table for inspection. This time, we used three linear fluorescent lights as a way to highlight defects. 【Software and Equipment Used】 Software used: DeepSky Field of view: Approximately 220 x 175 mm Minimum size of inspection target: 2 mm Number of inspection points: 1 Camera resolution: 1.3 million pixels Lens focal length: 12 mm Distance between lens and product: Approximately 400 mm Lighting: Three linear fluorescent lights Distance between lighting and inspection item: One light was set to illuminate from a diagonal direction (approximately 30°) to ensure that light hits the sloped surface of the inspection item.

• Image Processing Software

This is support after the introduction of DeepSky, a manufacturer of sheet metal parts for transport machinery and construction machinery. DeepSky can be used for various parts and defects, but the detection accuracy may vary depending on the settings. Although it is a one-time purchase inspection software, we will continue to provide support after the introduction. This time, we received a consultation regarding difficulties with learning. 【Inspection Settings and Results】 We received information about the varieties that were not working well and removed the "OK" label from the annotations. We also adjusted the annotation frames to only cover the defective areas and labeled them with two types: "NG" and "NG Crack" for learning purposes. Since the annotation frames were relatively large, I suspect that during learning, the AI might have recognized it as "there is an NG item roughly in this position on the left" rather than "looking at the defect," which may have caused confusion when NG items appeared on the right side as well. The size of the annotation frames is a significant factor affecting the accuracy of the defect positions in the teacher images. 【Software Used】 Software used: DeepSky

• Image Processing Software

This is an inquiry from a manufacturer of special steel parts such as automotive components. We have decided to conduct a simple verification regarding cracks in the riveting area. In visual inspections conducted by workers, there is often variability in the criteria for determining "good" and "defective" products among inspectors, and the more ambiguous the judgment based on severity, the more inconsistent the inspection results tend to be. We propose stable and efficient inspections using image analysis. 【Inspection Settings and Results】 It was determined that inspecting cracks on the outer circumference is difficult, as they cannot be captured from above, and side imaging tends to misdetect shadows from the uneven shape. In the range we verified, diagonal surfaces cast shadows when imaged from above, making it impossible to capture the cracks themselves. Detection from the side is also unstable, making inspection of these diagonal surfaces challenging. However, cracks on the upper surface that can be captured from above were detectable. Additionally, the presence or absence of burrs was also detectable from above. Our company website offers a free trial of our inspection software online. You can actually experience inspection software that uses AI (deep learning).

• Image Processing Software

This is an experiment to see if it is possible to distinguish a washer by annotating just one M5 washer in a single image and conducting 100 steps of learning with that one image. The left image shows the annotation (the task of enclosing to help remember what to find), while the right image places the M5 washer in the center with washers of different sizes on either side for verification. 【Inspection Settings and Results】 The washer could be found regardless of where it was placed on the screen. Even when multiple washers were placed and the lighting changed slightly, it was still able to identify them. The orientation did not matter, and for items with little variation, it could distinguish them with minimal learning. It recognized washers of different sizes and spring washers as washers. Nuts (with some shape variations) were also distinguished as washers. Items like mints, clips, and rollers in stock were identified as not being washers. It seems that items that are too large are not considered washers, while smaller items were recognized as washers. When an M5 washer was placed in between, it was clear that even though only one M5 washer had been taught, it was still able to make proper comparisons.

• Image Processing Software

We identified the front and back of washers based on requests from manufacturers of high-performance automotive parts and precision machined products. Since we had several types of washers in stock, we conducted a simple inspection using a model we prepared instead of the samples sent by the customer. 【Inspection Setup and Results】 Upon checking with the washers we have, we found it possible to recognize those with and without a taper as different items, so we believe that the inspection you inquired about is likely feasible. In the images, we used our software called DeepSky to train the system, designating the side with the taper as "Front" and the flat side as "Back" for judgment. To conduct a more detailed verification, there are generally two methods available: (1) Providing images that can be inspected on the customer’s side: If you can provide images taken directly from above the inspection item with a fixed distance from the camera lens to the item, we can consider whether we can perform the inspection. (2) Using a loaned device for verification on the customer’s side: It is also possible for you to directly assess the inspection feasibility using our loaned equipment. Depending on the availability of the equipment, you may need to wait your turn.

• Image Processing Software

This is a request for a simple verification from a manufacturer we have been working with for some time, who specializes in high-quality casting, machining, plastic processing, and the manufacturing and sales of acoustic products, delivering both domestically and internationally. This time, the focus is on cracks in the parts, and we attempted to analyze the provided images. Regarding "small cracks," we have been able to secure a sufficient number of samples, allowing us to hypothesize reasons for those that were not detected. However, for "large cracks" and "arm cracks," the limited number of samples resulted in insufficient findings in this verification. The report primarily features images of "small cracks," while other defects are included as reference images. Elements that are smaller, thinner, shorter, or darker than the cracks we were able to detect in this verification have not been identified. For "small, thin, and short" cracks, it is highly likely that they were not detected due to their size being too small relative to the overall field of view, and there is a possibility that dividing the field of view could allow for detection at the current resolution. For "thin and dark" cracks, it may be necessary to reconsider the imaging method. Additionally, if there are fundamentally difficult-to-detect types of defects, such as shallow cracks, the difficulty of detection increases significantly.

• Image Processing Software

Until now, we have posted several inspection cases of circuit boards using EasyInspector, but for inspections that determine "NG" if a solder ball is found anywhere on the entire board, we recommend DeepSky. DeepSky excels at inspections that classify a finding as defective if even one is discovered anywhere on the screen. 【Inspection Settings and Results】 By using the inspection function of "DeepSky," we were able to detect solder balls from the entire area and determine the inspection of one screen in 0.39 seconds. The work is displayed across the entire screen, and the task of finding the defective solder balls is referred to as "annotation," which involves surrounding the area with a frame. When the annotation is successful, parameters can be set for good judgment. To allow you to "experience" how much easier it is to set up compared to previous inspection software, the AI-based software DeepSky can be tried out on our company website. 【Software and Equipment Used】 Software used: DeepSky Field of view: 40x 30mm Minimum size of inspection target: 1mm Number of inspection points: 1 location, entire area Camera resolution: 1.3 million pixels Lens focal length: 25mm Distance between lens and product: approximately 200mm

• Image Processing Software

At the request of a battery charging and discharging equipment manufacturer we have been dealing with for some time, it is important in image inspection to determine whether we can clearly capture the defects such as "scratches" and "scuffs" that we want to identify. 【Inspection Settings and Results】 We rotated the received images by 60 degrees each, resulting in a total of 36 training images. Based on the images we processed, the scratches that were difficult to see were successfully detected using the images taken with the lighting you provided. If we cannot conduct verification with sample products, we cannot determine whether the defects we want to detect can be captured in the image inspection. We will ask you to send images where defects have been captured using various lighting techniques for evaluation. As expected, in the case of free evaluations, due to the fact that detection was performed on specific defective sample images and our imaging environment, we often guide users to actually experience it through the "demo machine trial service." 【Software Used】 Software Used: DeepSky Learning Version Number of Inspection Points: 1 (to find scratches and scuffs from the entire screen)

• Image Processing Software

We received a contact from a manufacturer of electronic components and devices regarding an issue where harnesses were slipping out of the binding process after wiring multiple harnesses with cable ties, causing difficulties. We suggested that it might be possible to conduct an inspection by counting the number of cable ties after receiving preliminary images, and subsequently, we actually took a sample for verification. [Inspection Setup and Results] In a simple verification, we were able to determine that the results were satisfactory. Regarding operations, we will consider the details while discussing the conditions in meetings. Given the current social situation, most meetings are held via web conference, but it is also possible for you to visit us directly. Many nearby manufacturers often bring samples, and since Hamamatsu City in Shizuoka Prefecture is an area with a thriving manufacturing industry, we have a track record of providing support to various industrial machinery manufacturers, control equipment manufacturers, and end users. We are also accepting inquiries for inspections that you may have previously given up on. [Software Used] Software Used: DeepSky Learning Edition Number of Inspection Points: 1

• Image Processing Software

This is an inquiry from a specialized trading company in automatic control. It involves determining whether crimping is present on harness cables with connectors, assessing the color order of about eight bundled colored wires, and confirming the tightening of screws at four locations on the circuit board. It seems that there are multiple types of cables, and this time we decided to utilize the strengths of both DeepSky and EasyInspector for the evaluation. 【Inspection Settings and Results】 Since the inspection items for insertion errors and defective crimp terminals are different, we verified each with separate software. For the cable insertion errors, we used EasyInspector to check for the presence of specified colors, while the inspection for defective crimp terminals was verified using the software DeepSky. Each software has its strengths and weaknesses. → EasyInspector struggles with inspections that have an infinite number of patterns. → DeepSky is not good at determining when one out of several colors is different. The left image shows the detection frame from DeepSky, and the right image shows the detection results from EasyInspector.

• Image Processing Software

A manufacturer of high-performance precision plastic molds and press molds has contacted us through our website to automate visual inspection. On our website, you can download the 60-day free trial software of EasyInspector and try out DeepSky's web version. Please check it out. 【Inspection Settings and Results】 We submitted the results for both EasyInspector and DeepSky. Considering the variability in the pad suction process and the uncertainty of where defects may appear on the parts, it is likely that DeepSky is more suitable for this situation. The web trial version of DeepSky only allows you to find what you have taught it. If you would like to verify the OK/NG judgment, we can lend you the full version for free. 【Software Used】 Software used: EasyInspector, DeepSky Learning Edition Number of inspection points: 1 location, finding defects from the entire screen.

• Image Processing Software

In recent years, the decrease in the labor population has become a social issue, and amidst this, the necessity for AI in inspections within the FA industry is increasing to maintain quality. Even among acoustic equipment manufacturers, our image inspection software is being considered. This time, we received a request for evaluation of polarity inspection for circuit board capacitors. 【Inspection Settings and Results】 A total of 23 images were used, including 3 images for NG evaluation, with a total of 11 training images consisting of 5 OK images, 5 NG images, and 1 evaluation image. Only the polarity markers of the capacitors were annotated, and after executing 12,000 steps of learning, judgments were possible. DeepSky is designed to include only the necessary functions for detection and judgment, and we plan to develop external applications for functions that may be required for customer-specific operations. 【Software Used】 Software Used: DeepSky Learning Version Number of Inspection Points: 8

• Image Processing Software

This is an inquiry from an electronic component manufacturer. Recently, there has been an increase in business locations where sending samples is difficult due to compliance and security concerns. DeepSky will create settings by learning from defective areas. Therefore, depending on the inspection content, if sending samples is difficult, it will be necessary to prepare about 20 to 30 images of good products and defective product data. First, it is essential to understand what kind of inspection content is involved, and it often starts with sending a few images of both acceptable and unacceptable products. [Inspection Settings and Results] As a result of the verification conducted with the images you provided, it was possible to distinguish between good and defective products. Since the images were 8-bit monochrome images, they were converted to 24-bit before verification. This verification was conducted using software called DeepSky, which utilizes AI (Deep Learning). By training the software to recognize the areas to be detected, it adjusts its own setting parameters and learns to recognize them. Good and defective products were trained with different labels to enable identification.

• Image Processing Software

This is an inquiry from a FA control equipment manufacturer. Visual inspection is often seen as a simple task, but it requires the skill to determine "OK" or "NG," and it may not be a task that everyone can perform consistently. Factors such as physical condition and prolonged working hours can lead to fatigue, which is one of the causes of unstable inspection results. Our company is developing inspection software that can identify fine scratches and dirt, enabling inspection judgments similar to those made by skilled workers. 【Inspection Settings and Results】 The device's claw mounting is being judged as good. The software allows for image data collection, learning, and setting of judgments in a manner similar to operating Windows software. It is an easy-to-integrate inspection software available for a one-time purchase. "DeepSky" offers a trial service online, so please check our website even for inspection cases you may have previously given up on. 【Software Used】 Software used: DeepSky

• Image Processing Software

We received an inquiry from a manufacturer of FA control devices with whom we have had dealings in the past, regarding the determination of solder quantity. The combination of AI (deep learning) and image inspection can be beneficial even in high-level solution areas. Please consider our image inspection software, DeepSky, for efficient inspection. 【Inspection Settings and Results】 We are able to accurately determine insufficient and excessive solder, as can be confirmed through images. With our inspection software "DeepSky," you can easily set up inspections by enclosing the areas you want to evaluate and training the system. There is an increasing movement to automate visual inspections due to advancements in AI-based image evaluation technology. To successfully implement AI image inspection, it is crucial to clearly define the characteristics of defects. Please feel free to contact us regarding lighting and camera environments as well. 【Software Used】 Software Used: DeepSky

• Image Processing Software

This is an inquiry from a manufacturer of FA control equipment with whom we have previously done business, regarding the evaluation of solder. The combination of AI (deep learning) and image inspection can be beneficial even in high-level solution areas. Our image inspection software, DeepSky, enables efficient inspection. 【Inspection Settings and Results】 We have determined the solder quality positively, as can be confirmed in the images. The numbers represent the AI confidence level percentage (number of recognition points). The inspection was conducted using software called DeepSky, which utilizes AI (deep learning). By training the software on the areas we want to detect, it adjusts its own setting parameters and learns to recognize them. To successfully implement AI image inspection, it is crucial to clarify the characteristics of defects. Please feel free to contact us regarding lighting and camera environments. 【Software Used】 Software used: DeepSky

• Image Processing Software

At a plastic parts manufacturer, counting small products for shipping has become a burden. There were issues with errors in weight inspection. We will conduct a simple verification of counting small parts. With DeepSky, you can use advanced software that supports counting on a conveyor. It is designed to count in various operational methods. 【Inspection Settings and Results】 We conducted inspections using software called DeepSky, which utilizes AI (Deep Learning). By training the software on the areas we want to detect, it adjusts its own setting parameters and learns to recognize them. The inspection cycle is shorter with EasyInspector. However, counting work that is not placed in trays is difficult to judge, so we decided to verify it with DeepSky this time. The left image shows the annotation, and the right image shows the detection frame. There were misjudgments with work that perfectly overlapped without any gaps, but otherwise, the judgments were good. 【Software Used】 Software used: DeepSky Counting can also be done with the application specialized for counting, 'cazoetell'.

• Image Processing Software

This is an inquiry from a manufacturer of automotive parts and accessories. I believe that clearly printing details such as model numbers and manufacturing dates on delivered products is a common practice across various industries. Our software includes basic functions for saving inspection results in CSV format or as images. It allows for the recording of which parts are associated with which product numbers, making it useful in various industries. 【Inspection Settings and Results】 By using the "OCRPro" feature of EasyInspector, we were able to read characters in 26 locations. The reading took 1.87 seconds. For characters, detection was possible using either the OCR function or the OCR Pro function, while for marks and logos, comparison inspection with master images was utilized. The images show differences in RGB and color tones in the "comparison with master image" settings. As for recording methods, results can be saved in CSV format or as images. Information like that in Table 2 will be recorded in CSV format.

• Image Processing Software

The interior of vehicles, such as cars, can be customized in terms of color and decoration according to the grade. We received an inquiry regarding color matching from an industrial machinery manufacturer with whom we have had previous dealings. They sent us sample images, and we conducted a simple verification. Our company website offers a free trial of our inspection software online. You can also experience our inspection software that uses AI (deep learning) firsthand. 【Inspection Settings and Results】 Out of 45 images, 15 (5 each) were used as training images. We inspected a total of 45 images, including the 15 training images and 30 untrained images. The result was that all 45 images were correctly detected. The breakdown of the images is 11 silver, 19 red, and 15 white. 【Software Used】 Software used: DeepSky

• Image Processing Software

This is an inquiry from a manufacturer of thermosetting and thermoplastic resin molding processing. The inspection is based on the images provided. 【Inspection Settings and Results】 The left image shows the annotation work to outline the area to be inspected. The right image shows the detection frame. There were no instances where defective products were incorrectly classified as acceptable, but there were 2 images where acceptable products were incorrectly classified as defective. Since September 2021, a data augmentation feature for training images has been implemented. By using this convenient feature, it is possible to learn from a larger number of images with variations in orientation, angle, and brightness of defects that occur only occasionally, which improves accuracy. 【Software Used】 Software Used: DeepSky Learning Version Number of Inspection Points: 4

• Image Processing Software

One of the frequently asked points from end users who are introducing image inspection for the first time is the ability to operate it in a manner similar to Windows software. This year, we have also developed related software that is useful for actual operations, such as the highly requested "simultaneous inspection and counting on a conveyor" and "OCR for difficult-to-read engravings." Additionally, we can add an optional "extended command" that enables a wide range of system operations. [Inspection Settings and Inspection Results] The image involves a task called "annotation," where the area to be detected is enclosed in a frame. By "training" the enclosed area, we were able to detect the target. Furthermore, as an image of actual operation, for example, when an abnormality is found while continuously photographing moving products, actions such as lighting a lamp, sounding a buzzer, or stopping the conveyor can be performed.

• Image Processing Software

Automotive parts and other resin-molded components may have variations in the presence or position of holes due to differences in specifications. However, in the case of similar products, accidents have occurred where similar but different items were shipped because they looked almost identical. [Inspection Settings and Results] By using the "Comparison with Master Image" feature of EasyInspector, we were able to detect the presence or absence of holes and positional differences in eight locations, allowing us to determine similar-looking items (different products) in less than one second. [Software and Equipment Used] Software Used: EasyInspector300 Field of View: 400 x 300mm Minimum Size of Inspection Target: 2mm Number of Inspection Points: 8 Camera Resolution: 1.3 Megapixels (Basler acA1300-75gc) Lens Focal Length: 6mm Distance Between Lens and Product: 300mm Lighting: Backlight The current 'EasyInspector2' color package can perform inspections using the "Comparison with Master Image" feature.

• Image Processing Software

Inquiry regarding scratch assessment for 200×100 mm square orange lens (plastic lens). There are requests for scratch assessment of plastic lenses, assessment of plastic molded product dents (depressions), assessment of plastic laser printing, and other appearance inspection requirements. In the initial stage, a free simple verification will be conducted, followed by guidance for paid verification. 【Inspection Settings and Results】 We verified whether scratches could be detected using the image inspection software EasyInspector. As a result, detection was possible, but false detections occurred in other areas depending on the type of scratch, making it impossible to detect "only defective scratches." With our product, the AI image inspection software "DeepSky," which uses AI (deep learning) capabilities, the AI can automatically adjust the setting parameters and recognize only the scratches as the target object. Therefore, it is recommended as it is less affected by changes in appearance due to lighting variations. 【Software Used】 Software used: EasyInspector710 Currently, EasyInspector2 also has AI capabilities and supports various inspections. We will propose solutions tailored to the inspection target.

• Image Processing Software

We received an inquiry from a manufacturer that produces a wide range of products, including vinyl chloride monomers, polymers, and special PVC resins, with whom we have had a business relationship for some time. They asked if our group company could distinguish the shapes of molded products (specifically, symmetrical shapes) and contacted us regarding this matter. 【Inspection Settings and Results】 We used a total of 14 images as training data, consisting of 7 images of the left side and 7 images of the right side. We enclosed the object we wanted to detect (in this case, the entire workpiece) in a frame and labeled them as "LH" and "RH" respectively (as shown in the left image). This process was carried out for all 14 training data images. In the right image, LH is correctly recognized as LH, and RH is recognized as RH. There are no issues with recognition even if the left and right are swapped. 【Software and Equipment Used】 Software Used: DeepSky Field of View: Approximately 482 x 383 mm Minimum Size of Inspection Target: 150 mm Number of Inspection Points: 1 (to check if the left and right workpieces are aligned across the entire screen) Camera Resolution: 1.3 million pixels Lens Focal Length: 8 mm Distance Between Lens and Product: Approximately 575 mm Lighting: Indoor fluorescent lights

• Image Processing Software

The samples you sent this time included a clear significant short and a subtle short that is not easily noticeable at first glance. It was confirmed that even minor shorts exhibited surface roughness associated with the short. With the inspection using EasyInspector, it would be easier to report if you could send us a "master image (good product)" and a "defective product that seems the most difficult to detect." 【Inspection Settings and Results】 Using EasyInspector's "color comparison inspection - comparison with master image" inspection function, we were able to conduct the inspection in 0.45 seconds. Even subtle shorts that were close in shape to the master image were detected for surface roughness (reflected white) and judged as unacceptable. The arrow markings in the photo are also white, but this time they were excluded from detection due to the non-detection pixel settings.

• Image Processing Software

There have been instances where defects such as scratches, bumps, and foreign matter contamination occur in resin-molded parts like automotive components due to processing methods, causing difficulties. 【Inspection Settings and Results】 By using the "Scratch and Bump Inspection" feature of EasyInspector, we were able to determine scratches and bumps across the entire field of view in 0.5 seconds. Two inspection frames were created for the entire area, setting them apart as "white" and "black." 【Software and Equipment Used】 Software Used: EasyInspector710 Field of View: 10 x 6 mm Minimum Size of Inspection Target: 2 mm Number of Inspection Points: 1 (entire area) Camera Resolution: 300,000 pixels Lens Focal Length: 50 mm + 10 mm Macro Ring Distance Between Lens and Product: 190 mm Lighting: Ring Lighting Distance from Lighting to Inspection Item: Not recorded The current 'EasyInspector2' color package can be used for [Scratch and Bump Detection].

• Image Processing Software

We received an inquiry from a manufacturer specializing in automatic control design and maintenance. They were considering conducting visual inspections for container lids made from injection-molded products, which are placed on an index table and processed at a rate of 18 pieces per minute (3 seconds per piece). They attempted to detect "scratches," "welds," and "spot debris." 【Inspection Settings and Results】 Initially, we conducted verification using images captured within the illuminated area. In the verification environment, it was possible to identify scratches, welds, and spot debris. Additionally, for weld lines specifically, an imaging method using a polarizing filter (as shown in the left diagram) proved effective. Our company offers verification and support from technical staff on a daily basis. If you have any issues, questions, or uncertainties during operation, please feel free to contact us at any time. 【Software Used】 Software Used: DeepSky

• Image Processing Software

We received a contact from a manufacturer that produces hardware and software for printed circuit board manufacturing and assembly. They are considering whether it is possible to perform image diagnostics to assess the quality of soldering after components are mounted on the printed circuit board and soldered. 【Inspection Settings and Results】 Based on the images provided, we were able to identify the NG (not good) areas as NG. In this verification, we used a total of 9 images: 6 images (NG1-3, OK1-3) as training data to create the model, but since we could not detect NG7-8, we added these 3 images to the training set. As a result, we were able to detect correctly, but since the occurrence of NG was fixed at the far right, further verification is needed to determine if detection is possible with other patterns.

• Image Processing Software

We received a request for a simple evaluation from a manufacturer specializing in wireless power supply and charging products, which we have previously contacted. Visual inspection is often seen as a simple task, but it requires the skill to determine "OK" or "NG," and it may not be a task that everyone can perform consistently. Factors such as physical condition and prolonged working hours can lead to fatigue, which is one of the causes of unstable inspection results. Our company is also developing inspection software that can identify fine scratches and dirt, enabling judgment similar to that of skilled workers. 【Inspection Settings and Results】 We tested with the 10 images provided. The images were assigned as follows: 1-3 → Test images, 4-6 → OK training data, N7-N10 → NG training data Although we could not secure enough NG test images due to the small number, we conducted training with 7 images and were able to correctly classify the 10 images. 【Software Used】 Software used: DeepSky

• Image Processing Software

We received an inquiry from a parts manufacturer considering the automation of various inspections for painted sheet metal and electronic components. In the initial simple inspection we will provide, we ask customers to prioritize their needs, and we will conduct free evaluations starting with the highest priority inspections. 【Inspection Setup and Results】 We conducted a simple evaluation to check for any mistakes in the insertion of the connector cables. As shown in the image, we were able to make a good judgment. Items that can be captured on the same screen can be inspected simultaneously. Depending on the PC specifications, it is generally possible to make judgments with an inspection cycle of about 0.2 to 0.3 seconds. In the free sample evaluation, we will first conduct a simple verification using the samples or images provided and report the verification results. In the simple verification, we will evaluate whether the desired detection/judgment is possible using our in-house equipment. If detection or judgment is possible in the simple verification, we recommend conducting a test (feasibility verification) assuming actual operation, where we will evaluate processing time and judgment accuracy. If you wish to conduct feasibility verification, you can choose to continue with our company (for a fee) or use our loaned equipment for verification at your company.

• Image Processing Software

In wire harness manufacturing, there is a process of bundling multiple harnesses together. While traditional "EasyInspector" is often used for inspecting electronic components, combining it with "DeepSky," which utilizes Deep Learning, can sometimes lead to higher quality production. 【Inspection Settings and Results】 Even with harnesses of various colors, we were able to accurately assess the bands regardless of their front or back. In the images provided this time, the large field of view made detection more challenging. We advised that taking close-up shots would improve accuracy. The numbers in the images represent the AI confidence percentage (number of recognition points). 【Software Used】 Software Used: DeepSky The current EasyInspector2 is equipped with AI capabilities and can handle inspections of defects with various colors and shapes. We will suggest software tailored to your inspection targets.

• Image Processing Software

We received an inquiry from a manufacturer of automotive and construction machinery harnesses. They would like to conduct inspections using our software based on a request from the client, as they cannot confirm the state of terminal crimping through image inspection. A free preliminary evaluation was conducted. In the free sample evaluation, we first perform a simple verification using the samples or images provided, and we report the verification results. The simple verification assesses whether the desired detection/judgment can be achieved using our internal equipment. If detection or judgment is possible in the simple verification, we recommend conducting a test (feasibility verification) assuming actual operation, and evaluating processing time and judgment accuracy. If feasibility verification is to be conducted, you can choose to continue with our company (for a fee) or use our loaned equipment for verification at your company. 【Inspection Settings and Results】 We conducted verification using our "DeepSky" equipped with deep learning capabilities. As a result, we were able to detect and judge each NG item. The left image shows detection frames indicating two types of abnormalities. The right image displays the number of detections that can be confirmed on the settings screen. However, in this verification, there were some instances of false judgments. Therefore, I believe it will be necessary to increase the number of verifications in the future.

• Image Processing Software

Manufacturers dealing with various types of motors have shown interest in our AI-based image inspection software "DeepSky" since its launch. This time, it will assess scratches and dents on screws. Our inspection software has over 2,000 examples of image inspections. We have numerous inspection cases published, including metals, plastics, food, electronic circuit boards, and pharmaceuticals, so please check our website to see if there are similar cases to the inspections you are considering. 【Inspection Settings and Results】 As a result of verification using the samples we received, we were able to detect scratches, allowing us to report favorable inspection results. We can respond free of charge to requests for demonstrations, web meetings, and demo unit loan services. The loan period for demo units is approximately one week, but availability is limited, so there may be delays. The image on the left shows annotations, while the image on the right shows the detection frame with a favorable assessment.

• Image Processing Software

We conducted a simple verification of dents and scratches on metal products such as motorcycle and automobile parts. Due to their almost identical appearance, there were incidents of shipping similar defective products, which caused issues. In the inspection process, automation is being promoted with the aim of pursuing stable inspection accuracy and cost reduction. As the shortage of personnel due to the aging workforce becomes more pronounced, we encourage you to consider improving operational efficiency through image inspection to protect "Made in Japan" manufacturing. 【Inspection Settings and Results】 Using the samples we have on hand for verification, we were able to detect dents and scratches. However, small dents and scratches present on good products, or surface conditions that appear as such in images, were detected as abnormalities. For extremely small items, we believe there are ways to address this by categorizing detection targets into "large dents" and "small dents," and only marking it as NG when a "large dent" is found.

• Image Processing Software

This is a simple verification using sample images at the request of an industrial machinery manufacturer. When introducing image inspection, the shooting environment for the inspection area becomes very important. Recently, there have been many cases where it is difficult to send sample products due to compliance and security reasons. While we can accept verification through image submission, there may be instances where we cannot provide guidance for a better imaging environment. In such cases, we may suggest creating a model in-house to conduct the verification, even though the texture of the sample may change. 【Inspection Settings and Results】 When learning with defective images that have significant differences compared to good products as training data, it was possible to detect those with large differences as defects. However, there were some cases where it was difficult to detect those with small differences. Regarding image acquisition, we determined that detection would be difficult if imaging was performed at a free distance for the current inspection item. If the imaging environment is adjusted so that the rivet appears in the center of the image at the same size, there is a possibility that detection accuracy could improve. 【Software Used】 Software used: DeepSky

• Image Processing Software

This is a specific consultation regarding the operation after a simple verification from a manufacturer of precision processed products such as automotive parts. We will create a setup to detect the blurred areas of a washer through image processing. In the free sample evaluation, we will first conduct a simple verification using the samples or images provided and report the verification results. During the simple verification, we will evaluate whether the desired detection/judgment is possible using our in-house equipment. If detection or judgment is successful in the simple verification, we recommend conducting a test (feasibility verification) assuming actual operation, where we will evaluate processing time and judgment accuracy. If feasibility verification is to be conducted, you can choose to continue with our company (for a fee) or verify using our loaned equipment at your company. [Inspection Setup and Results] Upon verification with the samples provided, it was possible to distinguish the presence or absence of blur. Taking images slightly diagonally rather than directly from the side showed better differences in appearance due to the presence of blur. We used 13 images as training data, framed the areas we wanted to inspect, and labeled all relevant areas of each image as "Blur Present" or "No Blur." We executed approximately 500 steps of learning and set it up so that it would be marked as a failure when a "No Blur" label was found.

• Image Processing Software

We received a request for a simple evaluation regarding the inspection of cracks in parts from an audio equipment manufacturer we have been working with for some time. We received 16 sample images. Our company accepts verification and support from technical staff on a daily basis. If you have any issues, questions, or uncertainties during operation, please feel free to contact us at any time. 【Inspection Settings and Results】 Out of the 16 images, 10 were used as training images. We inspected a total of 16 images, consisting of 10 training images and 6 untrained images (3 OK and 3 NG). We set the criteria to "pass" when there are 0 cracks detected and "fail" when 1 or more are detected, and we were able to detect all 16 images. Even with unknown images that were not trained, we were able to detect them as cracks. The image on the left shows the number of detected cracks displayed on the screen. 【Software Used】 Software used: DeepSky

• Image Processing Software

We received a request for a simple verification from a major general electronics manufacturer that has contacted us before regarding DeepSky. The presence or absence of washers is an area where DeepSky excels. We recommend trying out the web version to experience AI (deep learning) more closely. 【Inspection Settings and Results】 From the images provided, we used 10 OK images and 10 NG images, totaling 20 images, as training data. We then evaluated 5 OK images and 5 NG images, totaling 10 images, from the remaining images as unknown data. In this verification, there were no false detections, and we were able to correctly classify all 30 images. 【Software Used】 Software used: DeepSky

• Image Processing Software