LED chip recognition and positioning algorithm based on visual servo

With the continuous development of the electronic manufacturing equipment industry, the efficiency and accuracy requirements of automated manufacturing equipment continue to increase. In the LED chip inspection and sorting equipment, a visual servo platform is required to identify and position the wafer chips, and the positioning effect quickly and accurately. It directly determines the efficiency and quality of chip inspection and sorting equipment. In this paper, A “two-step positioning” method combined with two image recognition algorithms, is proposed to improve the positioning speed and accuracy of the LED equipment, which is fast coarse positioning and then slow precise positioning. In the first stage of rapid coarse positioning, the strategy of reducing the search angle and the search area is used, and on this basis, an improved shape matching algorithm is proposed to improve Recognition speed; the second stage fine positioning uses a recognition algorithm based on improved sub-pixel edge detection, which improves the recognition accuracy by improving the adaptive Canny operator. Experiments show that the two algorithms proposed in this paper combined with the “two-step positioning” strategy of LED chips can effectively improve the positioning speed and accuracy of LED chips to meet the needs of manufacturing.