AL will start from various viewpoints to find the parts with the largest difference in color, brightness, area, shape, edge, etc. between the registered OK product and the NG product, and automatically perform the most suitable detection settings, which can greatly enhance the defects in the above picture. Detection.
V200The first to introduce a defect detection algorithm based on CNN (Convolutional Neural Network) deep learning in the field of MLCC detection. It can still maintain a high detection capability under complex conditions such as component color changes. Frequent parameter adjustments are not required manually, greatly saving operation man-hours, and at the same time, it can effectively avoid the situation of inconsistent detection results caused by differences in parameters set by different personnel.