Precision positioning method of mini-LED chip based on YOLO algorithm

Abstract

The invention discloses a precision positioning method of a mini-LED chip based on a YOLO algorithm, and the method comprises the steps: setting a mark picture in a to-be-tracked mini-LED chip, and setting a preset pixel position; starting a camera after a conveyor belt is started; taking the mark picture as a template and storing the mark picture into a template file; performing target detectionon each frame of image from a first frame of image to obtain the confidence and the pixel position of the mark picture in each frame of image; determining an image matching condition according to a size relationship between the confidence of the image and a set confidence threshold; comparing the pixel position of the mark picture in the detected frame image with a preset pixel position, and adjusting the speed of the conveyor belt; and when the distance difference between the real-time pixel position of the mark picture and the pixel preset position is smaller than or equal to the allowable error, stopping the conveyor belt to finish positioning. According to the method, a YOLO deep learning target detection algorithm is adopted, the accurate positioning effect is achieved by tracking theposition of the mini-LED chip in real time and adjusting the speed of the conveyor belt in real time, and the production efficiency of the mini-LED is effectively improved.

Description

technical field [0001] The invention relates to the technical field of target tracking and positioning, in particular to a method for precise positioning of mini-LED chips based on the YOLO algorithm. Background technique [0002] At present, with the development of computer vision technology, vision-based moving target detection and tracking has become a current research hotspot, and has broad application prospects in video surveillance, virtual reality, human-computer interaction, planetary detection, and precision guidance. [0003] The YOLO algorithm is an object detection algorithm based on the deep learning neural network framework. The algorithm divides the input image into a 7 by 7 grid, and presets 5 default frames centered on each grid. The output of the algorithm is Predict the offset based on the 5 default boxes at each grid, and predict the corresponding category at the same time. The 5 preset default boxes are 5 representative boxes obtained by clustering on a...

AI Technical Summary

Problems solved by technology

At present, many machines use the method of template matching to realize the automatic positioning of chips in the process of automatic feeding of LEDs. There is a problem with the traditional method of using template matching. When the environment changes, such as light intensity changes, when the height of the terminal actuator changes , the camera needs to be re-calibrated, otherwise it will lead to serious inaccuracy of matching and low reliability, but this is very troublesome

Images