Research on Ampoule Injection Impurity Detection Algorithms Based on Background Estimation

It is not easy to autonomously detect glass crumbs and solid impurities of ampoule injection which still settle in the bottom of the bottle. In order to automatically identify impurities by detecting moving target in the bottle, ampoules are rotated at high-speed. Then it is stopped rapidly from motion. At that time, camera grab a series of images by for post processing. This algorithm makes use of background removal method to detect moving target. This paper researches the background estimated moving target detection algorithm base on Kalman filtering in the HALCON platform. It fits the trajectory of moving target by using point track-track matching algorithm which can solve the problem of multi-value associated. Impurities' characteristics of motion trajectory are different than bubbles'. In order to distinguish impurities and bubbles, detection, tracking and identification of impurities in ampoule injection are achieved. Simulation result shows that the processing speed of the algorithm is fast enough to be applied in the online system for detection of impurities in ampoule injection.