It is reported that the Micro LED automatic optical inspection system has the disadvantages of difficult feature extraction, low focusing accuracy, and slow detection speed. In order to solve these problems, the Jingcai Optoelectronics team independently developed Micro LED optical inspection equipment based on core technologies such as the LBG-YOLO deep learning algorithm of the automatic microscopic visual inspection system, the high-precision, large linear range autofocus method of the rectangular amplitude mask, and the neural network of the self-attention mechanism.
This device solves the pain point of difficult feature extraction, increases the detection speed from 78 fps of the standard YOLO v5 algorithm to 91 fps, and realizes the integration of Micro LED detection equipment.

ANNA