Optimization of image segmentation based on a visual positioning system for dispensing machines

Although dispensing technology plays an important part in the chip packaging process, most dispensing methods are inefficient. A visual positioning system for dispensing machines will greatly improve the efficiency. The key to visual positioning is an image processing technique that, among other things, includes image preprocessing and image segmentation. In this paper, we use the Otsu algorithm to provide segmentation of the original image and then apply the particle swarm optimization (PSO) algorithm to optimize the speed and accuracy of the image. By comparing the result with the exhaustion threshold segmentation algorithm we obtain the segmentation time and threshold, which shows that the PSO algorithm significantly improves the image processing speed.