The Research On Inspection Robot Of Ampoule Visible Particles

With the development of technology and the improvement of standard of living, higher requirements are needed by people in both quantity and quality of products. The traditional detection method of visible particle is carried out by man, but this method can not satisfy the requirement of high-speed and high-accuracy. A visual inspection system for ampoule is proposed in this paper, which achieves ampoule’s full-automatic inspection. This machine improves the speed and accuracy of inspection production line, which has great effect on economy and society.First of all, the requirement of visible particles detection in medicine and beverage is analyzed, proving the significance of the researching and developing the visible detection system. The definition of machine vision and its application in every field are summarized. Additionally, the present situation and trend of development of machine vision are researched.Secondly, the ampoule inspection robot’s system design method is put up. Besides, the production line structure, camera capturing method is analyzed. The structure and work flow of the ampoule inspection robot are introduced and the critical components are researched in detail in this article. The visible particle in ampoule is analyzed in this article. The research is taken on device chosen and platform established of vision inspection system covers aspects of illuminant technology, imaging method, optics lens, camera, etc.Furthermore, this article analyzes the inspection algorithm for visible particle. Image preprocessing, edge detection, level detection, moving object detection and other algorithms are researched. Based on the analysis of some favorite filtering algorithms, a self-adaptation weighted mean filtering algorithm based on mid-value is proposed in this article. SUSAN edge detection operator is used for image edge detection. Hough transformation is used for level detection. The improved method of frame difference is used for visible particle detection of ampoule. Sampling test is used for proving the effectiveness and feasibility of the detection method proposed by this article.In the end, the impact of performance of machine vision detection system from software design is analyzed. Several software design methods in machine vision system is summarized. The software interface of vision detection system in this article is introduced.