Last Updated： Aggregation Period:Feb 11, 2026～Mar 10, 2026
This ranking is based on the number of page views on our site.
Last Updated： Aggregation Period:Feb 11, 2026～Mar 10, 2026
This ranking is based on the number of page views on our site.
Last Updated： Aggregation Period:Feb 11, 2026～Mar 10, 2026
This ranking is based on the number of page views on our site.
1～30 item / All 466 items
Perfect for replacing visual inspections! Support all the way to environmental setup!
We provide appearance inspection software utilizing AI image recognition technology. It is capable of inspecting a wide variety of items, including metal processed products, films, resin molded products, and printed materials. (Examples: scratches, foreign objects, chipping, color unevenness, missing characters, etc.)
Fully automated AI appearance inspection combining object detection and robotics.
This is an appearance inspection software that utilizes AI object detection technology. It is capable of inspecting the appearance of a wide variety of workpieces, including metal processed products, films, resin molded products, and printed materials. (Examples: scratches, foreign objects, chips, color unevenness, missing characters, etc.) By combining it with industrial robots, it achieves a reduction in manual visual inspections.
A case study of automating visual inspection in production lines for continuous materials such as extrusion molding, hoses, and cables!
Regarding the "black spots and foreign object inspection" of extruded products, would you consider switching from "visual inspection" to "automated inspection using cameras" on the production line? We would like to introduce an example where installing cameras and lighting at the exit of the molding machine enables visual inspection of extruded products. By conducting inspections immediately after molding, it becomes possible to respond quickly if there are any issues with the molding process. 【Features of this system】 ■ Automated inspection can be performed without the need to introduce separate lines or equipment for inspection. ■ Since inspections are conducted immediately after production, rapid responses are possible in case of defects. ■ By arranging cameras to surround the product, comprehensive inspection is possible. ■ The equipment can be modified depending on the inspection requirements. *First, we will verify whether the work you wish to inspect can be accommodated. *Detailed processing information, including the processing flow, is provided in the PDF.
A case study on the automation of visual inspection to detect missing components, variations in types, and printing defects on components on the substrate!
Shall we switch the inspection of the presence or absence of components on the electronic circuit board and printing inspection from "visual inspection" to "automated inspection using a camera"? We would like to introduce a practical example of the "iVision Robot," which operates as an inspection device by attaching an inspection system to a general-purpose XY stage moving device. By using a two-dimensional operating stage, it becomes possible to check the mounting status of the entire circuit board, which cannot be confirmed by a single camera. 【Features of this system】 - It can automatically inspect the presence or absence of components and installation errors on the entire circuit board. - It can accommodate different types by setting up specifications for various products. - It is also possible to simultaneously inspect the status of the circuit. *First, we will verify whether the work you wish to inspect is feasible. *Detailed processing information and processing flow are provided in the PDF.
Would you like to automate simple visual inspections by installing a general-purpose image sensor on the conveyor line?
Are you worried about forgetting to attach parts to products flowing on the conveyor line or not having processing holes, and therefore conducting visual inspections, but finding that the cost of introducing automatic inspection equipment is too high? "The inspection itself is not difficult, but it becomes a problem if there is a leak, so we need to assign inspectors. However, we cannot spend nearly 10 million yen on inspection equipment for automation." If you feel this way, please consider us [we are currently offering a guide]. ADDI Co., Ltd. can provide delivery solutions tailored to your needs thanks to our long-standing expertise in the development of inspection equipment. Additionally, our staff will handle installation and initial setup, reducing the burden on our customers. "The inspection content is simple, so we can't spend that much," or "We don't have staff who can handle in-house production." The image processing visual inspection system "iVision" can solve these concerns! *For more details, please download the catalog in PDF format. We will first confirm with you "what" and "in what condition" you want to conduct inspections, consider the appropriate cameras and lighting, and then propose the necessary inspection equipment based on that.
![[Data] Case Studies of Appearance Image Inspection](https://image.mono.ipros.com/public/product/image/1a6/2000395120/IPROS79916081570846414894.jpeg?w=280&h=280)
Numerous image inspection case studies using the AI general-purpose appearance inspection software "EasyInspector2" are featured!
This document presents case studies of image inspections using the AI general-purpose appearance inspection software "EasyInspector2." Customers from the automotive parts and electronic circuit board sectors make up half of our clientele, and we serve a wide range of industries, including assembly verification of equipment and inspection of minute dirt on products. [Featured Case Studies (Excerpt)] - Inspection of defects in circuit board patterns and silk printing - Inspection of the orientation of electronic components - Counting of parts and detection of different items - Inspection of scratches on uneven surfaces - Measurement of dimensions and angles of mechanical parts - Counting the number of sheets of paper - OCR reading and barcode scanning - Reading of laser engravings on IC components - Meter display reading - Inspection of scratches and dents on aluminum die-cast parts - Hair contamination in food - Detection of several types of chemicals - OCR for metal engravings and more. *For more details, please refer to the PDF document or feel free to contact us.
Image inspection becomes dramatically easier with AI (deep learning)!
"DeepSky" is an AI visual inspection software that is easy for on-site users to operate with its "all-in-one & simple operation" design. It incorporates object recognition and good product learning from various deep learning methods, and further researches models optimized for factory automation (FA) to create its lineup. Skylogic, which has provided and supported over 2,000 examples of image processing products for FA sites, designed it with a focus on the needs of the field. ~ Do you have any of the following requests or concerns on-site? ~ ■ I want to start utilizing AI on a small scale. ■ I am worried about the judgment accuracy of the current inspection machine. ■ There are limited people who can set up the inspection machine. ■ It is difficult to secure and train inspection personnel. ■ There is variability in accuracy and inspection standards. DeepSky can perform "image data collection → learning → judgment" with just one PC. AI image processing is different from traditional procedural (rule-based) image processing in that it is easy to set up and excels in judgment under complex and variable conditions. We offer sample verification and a web trial service, so please feel free to reach out to us.
![[AI Image Inspection Case] Detection of Dirt and Scratches on Metal Parts](https://image.mono.ipros.com/public/product/image/123/2000669378/IPROS02005936974709862407.png?w=280&h=280)
Detects defects such as scratches, dirt, and black spots on the equipment.
The manufacturer of specialized processing jigs for building materials that we are inquiring about this time is a company that has requested numerous verifications from us, specifically to detect defects such as dirt and scratches on their equipment. We conducted the verification using the inspection function of "DeepSky." This time, we were able to successfully distinguish the scratches, dirt, and black spots. The software allows for easy collection of image data, learning, and judgment using just one PC, and we offer a web trial service that allows for convenient testing of detection performance, as well as videos showcasing numerous detection examples. Even for inspection cases that you may have previously given up on, we encourage you to check our website. [Software Used] ■DeepSky *For more details, please refer to the PDF materials or feel free to contact us.
![[AI Image Inspection Case] Transparent Plastic Defects](https://image.mono.ipros.com/public/product/image/30b/2000669379/IPROS09154910826230777000.png?w=280&h=280)
We will introduce a case where verification was conducted using the inspection software "DeepSky" and "EasyInspector."
This is a request from a manufacturer that produces plastic molded products. They have a cap with a diameter of approximately 30mm and a height of about 8mm, but sometimes black spots (approximately 0.2mm in diameter) appear on the flat top. They would like to conduct an appearance inspection of this cap and inquired whether it is possible to detect these black spots through their website. We conducted verification using the inspection software "DeepSky" (which utilizes AI and deep learning) and "EasyInspector" (a traditional rule-based inspection software). In this case, it can be said that verifying with 'DeepSky' is smoother. This software can be easily operated in a manner similar to operating Windows software, which leads to many direct inquiries from end-users who are introducing image inspection for the first time. [Software Used] ■DeepSky ■EasyInspector (formerly EasyInspector) *For more details, please refer to the PDF document or feel free to contact us.
![[AI Image Inspection Case] Foreign Objects and Defects on Non-Woven Fabric Masks](https://image.mono.ipros.com/public/product/image/164/2001402023/IPROS50812371807987008330.png?w=280&h=280)
Detect foreign objects on products processed with patterned non-woven fabric and masks.
There are three main categories of inquiries we receive regarding non-woven fabric inspections. - Inspection in a roll-to-roll state - Inspection of non-woven fabric with printed patterns - Inspection of non-woven fabric processed into products like masks The systems used vary depending on each state. For example, inspections in a roll-to-roll state often utilize EasyInspector, while detecting foreign substances on products with printed patterns or processed into masks is a specialty of DeepSky. Processed products have irregularities and overlaps, which can lead to frequent false detections with standard binarization methods. However, deep learning can learn and detect features such as insects, hair, and debris, making it largely unaffected by the variations in product patterns and processing. Here, we will introduce detections made by DeepSky. - Foreign substances such as hair and debris (Images 1 and 2) - Defects such as wrinkles and tears (Images 3 and 4) In roll-to-roll inspections, a method that continuously monitors with a fixed camera while the product is moving is suitable. However, in cases like the masks that have been processed into individual products, considering automatic discharge, the following system can be utilized.
![[AI Image Inspection Case] Type Identification of Metal Coated Workpieces](https://image.mono.ipros.com/public/product/image/206/2001402034/IPROS31089609427168558532.png?w=280&h=280)
It is possible to make judgments using DeepSky's inspection features! It is applied to various inspections.
We received two samples of metal coatings for truck and automotive parts from a customer engaged in low-volume, high-variety production. They were experiencing issues with shipping different products due to multiple part numbers flowing through the same line. Here, we present a case study on how we conducted verification to distinguish between the two samples. Using DeepSky's inspection capabilities, we were able to make determinations. DeepSky is designed solely for inspection functions, and by integrating it with PLCs and socket communication, it can be applied to various inspections. We have made commands available for input and output, allowing customers to create their own software for free. Additionally, we can also undertake software development, but this will be a paid service. 【Software and Equipment Used (Excerpt)】 ■ Software Used: DeepSky ■ Field of View: Approximately 447×335mm ■ Minimum Size of Inspection Target: 5mm ■ Number of Inspection Points: 1 ■ Camera Resolution: 1.3 Megapixels *For more details, please refer to the related links or feel free to contact us.
![[AI Image Inspection Case] Meter Reading](https://image.mono.ipros.com/public/product/image/43d/2001402035/IPROS04380832560386695565.png?w=280&h=280)
The lamp can detect and determine the difference in color between when it is on and when it is off!
We would like to introduce a case that was inquired about by an industrial equipment system manufacturer with whom we have previously done business. When the red light is on, it triggers an alert; the panel meter sends an alert at 10% below the set value, and the analog meter has a system that triggers alerts based on their respective settings. They were considering a setup where a camera is connected via wired LAN to a computer, which in turn is connected to a Patlite to sound an alarm. By using the "Meter Reading" function of EasyInspector, we were able to make determinations at two locations in 0.35 seconds. 【Software Used】 ■EasyInspector The current 'EasyInspector2' control panel and ControlPanel package [Meter Display Reading] can be inspected with 'EasyMonitoring2'. *For more details, please refer to the related links or feel free to contact us.
![[AI Image Inspection Case] Simultaneous Inspection of Defects and Runs in Metal Coatings](https://image.mono.ipros.com/public/product/image/e69/2001402052/IPROS60391880557747765672.png?w=280&h=280)
It is possible to set up to detect only those with high recognition scores or to display all with low scores!
We would like to introduce a case study of a simple verification of metal painted products for defects and liquid drips. Verification was conducted using DeepSky's inspection function, and detection was possible. As shown in the image, detection is taking place. The numbers represent the AI's confidence percentage, referred to as the number of recognition points. In DeepSky, you can set it to detect only those with a high number of recognition points as defects, or you can set it to display all items with a low number (anything that the AI perceives as somewhat similar to the defects it has learned). 【Software Used】 ■DeepSky *For more details, please refer to the related links or feel free to contact us.
![[AI Image Inspection Case] Presence or Absence of Indelible Stamp Characters](https://image.mono.ipros.com/public/product/image/6e2/2001412094/IPROS3372393238205387798.png?w=280&h=280)
We will check if the stamp with indelible ink of letters and numbers is applied!
We received inquiries about our inspection software from manufacturers of plastic molded products, including injection molding, processing, assembly, and mold production. We conduct a simple verification to check whether a stamp with indelible ink of letters and numbers has been applied. By using the "Presence of Specified Color Inspection" feature of EasyInspector, we were able to inspect two locations in 0.4 seconds. The left image shows the inspection frame set to specify the inspection locations. The right image displays the confirmed numbers as pass (in blue) and the letters without printing as fail (in red). 【Software and Equipment Used】 Software Used: EasyInspector (formerly EasyInspector) Field of View: Approximately 223 x 177mm Minimum Size of Inspection Target: 10mm Number of Inspection Points: 1 Camera Resolution: 1.3 Megapixels Lens Focal Length: 16mm Distance Between Lens and Product: Approximately 540mm Lighting: Ring Lighting Distance from Lighting to Inspection Item: Illuminated from 100mm above Current 'EasyInspector2' color package can perform inspections for the "Presence of Specified Color."
![[AI Image Inspection Case] Inspection and Recording of Product Presence After Packaging Work](https://image.mono.ipros.com/public/product/image/a70/2001415492/IPROS12682663565000975305.png?w=280&h=280)
After packing the products in the warehouse, we will inspect them for any shortages and record the results in a CSV!
We have received inquiries about our inspection software from shipping companies as well. After packing products in the warehouse, they are considering incorporating an image recording process into the workflow to leave evidence that there are no missing products. Our inspection software is being utilized in various industries. By using the "Presence of Specified Color Inspection" feature of EasyInspector, we were able to determine the pass/fail status of the number of items divided into six areas in 0.39 seconds. If the condition is normal, the detection of the specified color above the reference value results in a "pass" judgment. The left image shows the settings screen, while the right image displays the inspection results table. As for the recording method, it is possible to save the results as images or record them in CSV format.
![[AI Image Inspection Case] Inspection of containers with cushioning materials for transportation.](https://image.mono.ipros.com/public/product/image/b2f/2001415496/IPROS4081292102060013901.png?w=280&h=280)
Determine abnormalities in 12 locations of cushioning material corrugated cardboard!
We consulted with another company with a budget of 500,000 yen, but they said that a precise inspection could not be done in just a few seconds. We received a request from a distribution and logistics manufacturer who wants to implement a process where, without needing precision, they can make a preliminary judgment in a few seconds on whether there is a "possibility of being NG" for specific items, which would then be checked by a person. This is regarding the inspection of containers with cushioning materials for transportation. By using EasyInspector's "comparison with master images" function, we were able to determine abnormalities in 12 locations of the cushioning cardboard in 0.25 seconds. The left image successfully detected and judged a misinstallation of the Santec foam component. The right image incorrectly judged the defective part. To ensure accurate judgment, it is important to enhance the imaging environment by adding lighting to make the image brighter, clearly darkening the shadowed areas, and adjusting the camera position to be further away and as vertical as possible to minimize hidden parts due to the partition board.
![[AI Image Inspection Case] 7-Segment Display Reading](https://image.mono.ipros.com/public/product/image/6f3/2001415497/IPROS4062314720179571701.png?w=280&h=280)
I will read the 7-segment display meter at the construction site!
Our inspection system is also in operation in the architecture and construction industry. We verified the reading of a 7-segment display meter. By using the "OCR" function of EasyInspector, we were able to determine 7 characters. The image itself was slightly blurred, and there were color inconsistencies in the image. When we extracted and processed a portion of the image that seemed easier to read, it was not processed very cleanly, but we were able to read it nonetheless. I believe that improving the imaging conditions would allow for clearer readings. [Software Used] EasyInspector710 (formerly EasyInspector) The current 'EasyInspector2' CP (Control Panel) package [Digital 7-segment display reading] can be inspected with 'EasyMonitoring2'.
![[AI Image Inspection Case] Presence or Absence of Manufacturing Date Printing on Cardboard](https://image.mono.ipros.com/public/product/image/02b/2001415504/IPROS16821116889892353560.png?w=280&h=280)
We will conduct a simple verification to see if the manufacturing date of the confectionery manufacturer is printed on the cardboard box!
The confectionery manufacturer in this inquiry is packaged in cardboard boxes, just like other industries. We will conduct a simple verification to check whether the manufacturing date is printed on the cardboard box. The EasyInspector software (traditional rule-based image inspection software) that you inquired about is primarily used by customers in the automotive parts and electronic circuit board sectors, making up half of our clientele. Additionally, it is utilized by a wide range of industries for tasks such as assembly verification and inspection of minor product dirt. We plan to use the EasyInspector feature called "Color Presence Inspection." This function allows us to specify an ink color and determine whether it falls within the specified range for a pass/fail judgment. In the left image, the specified color is detected as red, and a passing blue frame is displayed. In the right image, the specified color could not be detected, resulting in a failing red frame. It is also possible to conduct simultaneous inspections at two locations by using two cameras.
![[AI Image Inspection Case] Counting Round Bars](https://image.mono.ipros.com/public/product/image/2e6/2001415506/IPROS9787694977205289228.png?w=280&h=280)
Counting the number of unpainted round bars using AI image inspection!
This is a request from a metal processing manufacturer to count the number of 7-meter long rods without positioning them. This time, the rods were unpainted, but they had a level of accuracy comparable to the previously tested white-painted rods. However, items with significantly different sizes or angles may be inaccurate due to lack of training. The left image shows the annotation, while the right image displays the detected work for counting as points. The strength of AI (deep learning) image inspection software over traditional rule-based image inspection software is its ability to perform inspections even with variations in brightness and imaging environments, such as outdoors. Inspection software can be beneficial across various industries and processes. At DeepSky, you can utilize advanced software that supports counting on conveyors. It is designed to allow counting through various operational methods. 【Software Used】 Deepsky
![[AI Image Inspection Case] Reading of Indicator Lights and Meters](https://image.mono.ipros.com/public/product/image/e86/2001415539/IPROS3785475091512456314.png?w=280&h=280)
I will read the indicators and meters of the control equipment!
We received inquiries from an electronic equipment manufacturer regarding the display lights and meter readings of various control devices. We have decided to set up and report on six types of reading functions. The two images on the left and right utilize the "meter reading function." For the other three images, we determine pass or fail based on whether the color is detected when lit or not. However, I believe there is variability in how the lighting hits the upper and lower parts of the meter, so I think that illuminating the lower part, which tends to be darker, with bar lighting or similar would reduce misdetections. Regarding the other image we received, reading was possible with the image you provided. Additionally, some minor adjustment for misalignment is possible, but fixing the camera is essential.
![[AI Image Inspection Case] Tears in Shrink-Wrapped Products](https://image.mono.ipros.com/public/product/image/2ab/2001415554/IPROS12208243345143742746.png?w=280&h=280)
Verify the presence or absence of tears in shrink-wrapped products!
We reported the presence or absence of tears in shrink-wrapped products from food manufacturers through a free simple verification process. We would like to continue verifying issues such as deformation, stacking collapse, and flap adhesion, and propose various defect detection methods. In the simple verification, we were able to accurately identify approximately 80% of tears that are easily noticeable by the human eye, and about 50% of tears that are less noticeable, such as those on white products. This time, we conducted the verification with a limited number of samples, and I believe the accuracy was affected by the small amount of training data. The detection accuracy will improve with more data for training. [Software Used] Deepsky
What is visual inspection? Points to consider for automation, the use of the latest AI, and trends overseas.
In the manufacturing industry, quality control and quality assurance are essential tasks for maintaining corporate trust. Among the various quality control tasks, "visual inspection" occupies a significant portion. By conducting visual inspections at each stage of the process, the yield of downstream processes can be improved, and the outflow of defects can be prevented in advance. However, this can also increase manufacturing costs and, in some cases, the visual inspection may become a bottleneck, leading to a decrease in production capacity. Here, we will discuss the points to consider for automating visual inspections, the differences between traditional image processing and the latest AI image processing, and trends in visual inspection overseas. ■ What is visual inspection? ■ Main inspection items of visual inspection ■ Automated visual inspection as an alternative to manual inspection ■ Advantages and disadvantages of manual inspection ■ Advantages and disadvantages of automated visual inspection ■ Steps for implementing automated visual inspection ■ Will AI (deep learning) visual inspection replace traditional rule-based image processing? ■ Situations where procedural image processing is used in visual inspection ■ Situations where AI is used in visual inspection ■ Points to consider when automating visual inspection ■ Trends in automated visual inspection overseas ■ Summary ■ Visual inspection systems that can be implemented at low cost *Please see the related links.
It is necessary to consider whether it is truly impossible to make a judgment without using a high-resolution camera when selecting the camera's pixel count.
"High pixels mean high performance" — Indeed, as the pixel count increases, it becomes possible to capture details more clearly. However, it does not simply mean that larger is better, as high-pixel cameras also have drawbacks. For example, using a camera with unnecessarily high pixel counts can lead to longer image processing times. Therefore, it is necessary to consider whether "it is really impossible to make a judgment without using a high-pixel camera?" when selecting the pixel count of a camera. The example below shows the results of an experiment on how much difference in clarity there is when enlarging a part of a PC motherboard from 300,000 pixels to 14 million pixels. (Refer to the figure) In this way, high pixel performance is demonstrated when trying to see fine details within a wide field of view. However, as mentioned earlier, not only does image processing take longer, but as will be introduced in "High pixels or low pixels (2)," it is also necessary to increase the exposure time. The consideration of whether "it is really impossible to make a judgment without using a high-pixel camera?" is one of the important factors in selecting a camera.
About dimming and lens aperture.
We received a question from a customer using the image inspection software EasyInspector, asking, "Can inverter fluorescent lights not be dimmed?" We inform customers using EasyInspector that "if there are no issues with inspections using indoor fluorescent lights or inverter fluorescent lights, there is no need to use expensive lighting." Many LED lights have separate lighting and power units, and since the power unit can be dimmed with a knob, I thought such questions were natural. In fact, there are various methods for adjusting brightness: 1) Dimming 2) Aperture of the lens (in the case of industrial cameras) 3) Camera exposure time and gain settings In other words, in principle, if any of the above three can be adjusted, brightness can be controlled. So, what is the best method to use for adjustment? (Figure 1) This is a photograph of part of a bottle cap. The focus is set on the bottom of the cap. Both images have the same brightness, but you can see that the left image has the focus extending to the edge of the cap compared to the right. This is referred to as "having a deep depth of field."
![[Knowledge of Visual Inspection] A discussion about what AI is doing, I wonder.](https://image.mono.ipros.com/public/product/image/380/2001417610/IPROS18132662272022808754.jpg?w=280&h=280)
If you have something you want to find, please feel free to contact Sky Logic!
When talking with customers, I sometimes feel that they believe AI (deep learning) automatically determines what is OK or NG. This is a subtle nuance, and it's not something that needs to be corrected in the course of actual conversation, but today I want to clear up that confusion by writing this article. Now, regarding the confusion mentioned above, the actual difference is that it is humans who decide what is considered OK and what is NG in the software. This leads to the question, "So what does AI do then...?" To put it simply, what our AI does is "find what it has been taught," and fundamentally, that’s all there is to it. This function is generally referred to as "object detection." In "object detection," it literally detects "objects" within an image.
Introducing a system created with DeepSky and external software that inspects products moving on a conveyor while counting them!
DeepSky has the capability to communicate the coordinates of detected objects to a higher-level system via TCP/IP, enabling various applications by tracking the coordinates of objects within that system. Here, we introduce software that inspects objects flowing on a conveyor and counts them when they cross the center of the camera. The diagram below (Figure 2) illustrates the configuration of the conveyor, camera, and software. The camera's images are processed by DeepSky to detect the type and coordinates of the objects, and this information is passed to the upper counting software. The counting software tracks the objects and increments the count when they cross the center of the image. This is an image of products moving on a belt conveyor. The upper camera captures images while checking for defective items and counting the products. It can also interact with PLCs to reduce speed or trigger a buzzer when approaching a specified count. The counting software tracks the coordinates of the objects to count them. In this way, DeepSky can work in conjunction with higher-level software, allowing for various uses. Custom software, like the counting software, can also be developed by our company.
I will explain the differences between "cazoeTell" and "wakeTell (formerly known as cazoeTell2)."
The name "cazoeTell2" has been changed to "wakeTell." The reason for the renaming is that the name "2" often leads customers to mistakenly think it is a superior version of cazoeTell. cazoeTell is specialized in counting a single item and has high counting performance. wakeTell can classify and count multiple items, but its counting performance is lower than that of cazoeTell. ■ Comparison Table (Please refer to the related link (blog)) ■ Comparison of Counting Nuts and Washers Both cazoeTell and wakeTell can count a total of "30" items without any issues. However, cazoeTell only provides the result that there are "30" items without specifying how many of each item there are. On the other hand, wakeTell recognizes and displays the result as "15" nuts and "15" washers. Therefore, wakeTell can also be used for purposes such as forgetting to return tools.
![[Technical Support] Would you like to try AI in food inspection?](https://image.mono.ipros.com/public/product/image/7d7/2001418491/IPROS14729595557968094275.jpg?w=280&h=280)
It can be used for food inspections such as material defects, foreign matter contamination, insufficient quantity, and missing components!
Until now, Skylogic has primarily handled image inspection projects focused on industrial products. This is because the nature of industrial products, which involves "mass-producing the same item," matched well with conventional (rule-based) image processing methods. Conversely, it has been difficult to handle items with unstable colors and shapes, even among industrial products, using traditional methods. However, with the emergence of image processing methods utilizing AI (deep learning), items with variability can now be included in the scope of image processing. This is because AI excels at "detecting only what we want to find while ignoring trivial changes." Now, connecting this to the title, when we talk about "items with unstable colors and shapes," we are referring to food products. Issues such as material defects, foreign object contamination, quantity defects, and missing components... even when the subject is food, products produced in factories are likely to face similar challenges as industrial products. However, the difficulty of processing with traditional methods has led to these issues being abandoned. (And I have also given up on them.) Why not try AI?
Skylogic has already developed AI for detecting foreign substances in food.
▼Obtaining Anisakis → Until Filming This time, we conducted a verification at our company to find "Anisakis," which we received several inquiries about from customers, using AI. We started by trying to obtain live Anisakis, but it turned out to be more difficult than we expected... We visited several supermarkets and fish shops in the city and made phone calls to request their help. Despite it being an extremely busy time of year, I made a strange request to unknown housewives, saying, "Can you give me live Anisakis for an experiment to find Anisakis with AI (to summarize)?" I am very grateful to the supermarkets that cooperated with us, and to the staff in the fresh fish section who helped us. Thank you very much. We assembled the equipment assuming an operation where we conduct AI Anisakis inspections on a slowly moving conveyor, and if detected, a buzzer would sound to stop the conveyor. Anisakis glows in response to specific wavelengths of ultraviolet light, so we use UV LED light sources and filters that pass specific wavelengths. In image processing, it is also important to capture images that make it easier to find Anisakis through optical manipulation.
I will explain the principles of object recognition!
Recently, we have been receiving more inquiries from customers using AI software (such as EasyInspector2 and DeepSky) asking questions like, "Why does this happen?", "Isn't it supposed to be like this?", and "What is going on inside?". In such cases, it seems that the person in charge is unsure about how to explain things. Indeed, when trying to provide an intuitive and easy-to-understand explanation, they often end up relying on analogies, or if they attempt to explain in detail, they find that specialized books provide more thorough information. I have also been unable to find suitable explanatory materials for customers who want to take a deeper dive into the principles of AI image processing.
