Final substrate appearance inspection device

Final substrate appearance inspection device with excellent defect identification capability powered by artificial intelligence.

In the conventional learning method used for final appearance inspection devices, as the number of training samples increases, the number of pseudo-defects decreases, but the defect detection capability drops drastically. Particularly, minute defects near patterns become almost invisible. Our company has successfully developed a new technology that overcomes the shortcomings of such learning methods. This technology changes the weights and connection states of the artificial intelligence parameters for defect detection through learning, allowing for accurate detection of defects near patterns, which were previously considered a drawback of conventional learning methods, while generating almost no pseudo-defects. With the development of this technology, it has become possible to detect defects such as solder mask peeling and excessive solder on solder substrates, which were previously deemed impossible to detect.

Visual Inspection EquipmentImage analysis software