A Positioning Optimization System for Intelligent Pharmaceutical Production Line Based on Visual Imaging
Yingkun Hu, Xiufang Zeng, Pinzhong Chen · 2023
On the pharmaceutical automation production line, the application of artificial intelligence technologies such as machine vision can make pharmaceutical production automated and intelligent, improve the quality and efficiency of pharmaceutical production. It can be said that machine vision is a major core technology for the automation and intelligence of pharmaceutical production equipment. This article focuses on the visual positioning technology for the filling and sealing process of traditional Chinese medicine in pharmaceutical automation production lines. The research content mainly includes two aspects: The design of visual imaging systems and the research on visual positioning algorithms for bottle mouths, stoppers, and caps. The visual imaging system uses an industrial camera combined with a circular light source, and then uses a hand eye calibration installation method to capture high-quality images of bottle mouth, cork, and cap on a movable robotic arm. Next, a series of image preprocessing operations are carried out on the original image, including Gaussian filtering and smoothing of the image to remove noise, gamma transform image enhancement and edge detection preprocessing operations, and the use of Canny edge detection to obtain images with good foreground and background contrast, appropriate brightness, and prominent edges. Finally, a circle fitting algorithm based on KMeans++ was used to locate the extracted edge images of bottle mouth, bottle stopper, and bottle cap, and the experimental results were preliminarily analyzed.