Research And Design Of Drinks After Filling Impurities Inspection System Based On Machine Vision System

In recent years, the state and consumers pay more and more attention to the safety of food, and beverage manufacturing enterprise also pay more and more attention to the beverage quality inspecting. With the improvement of the automation degree in filling lines, artificially detection method which is used by many enterprises can not meet high speed line inspection requirements, especially at present,in drinks after filling inspection field, automation level is still very low, which is the development bottleneck of automatic drinks filling lines in our country. So we need a suite of online after filling inspection equipment to improve the whole line inspection level. Machine vision technology is fast and has accurate inspection effect, and it has proved its superiority in many fields. The purpose of this project is to research a method which adopts machine vision technology to complete drinks after filling impurity inspection to satisfy the present requirements.This subject design a drinks after filling impurities inspection system using image processing method,according to the actual process and inspection requirement. Its main research contents are as follows:1. We confirm the image acquisition plan. First we introduced the image collection equipment and the basis of choice, then according to the features of inspection target, we select a set of suitable image processing equipments, including light sources and lighting system, cameras and lenses selection and selection of image processing equipment.2. We confirm the main workflow of the system according to current artificially testing process. Then we choose to use general software development tools and C/C + + language to complete algorithm design. In the end, After analyzing the structure of the image processing algorithm, the time complexity of image processing algorithm and the parallel processing method which is commonly used now, we choose cooperative work with multiple processors means to complete the image processing algorithm. We also distribute multiprocessor's own image processing tasks, at last we explained the content of information communication among multiprocessor.3. We confirm the system's hardware structure on the basis of the function desire of the online inspection system combining with the structure design scheme of the image processing algorithm. We design the structure figure of the system's hardware, and insured the connection mode among the image processing equipment.4. According to the features of target images, we determine the impurities inspection system, including localization algorithm of interested region, background suppression, image smooth, image enhancement, image segmentation algorithm, connected domain analysis and correction, connected domain feature extraction, connected domain feature matching, target motion trajectory generation and target properties judgment. We complete code design in Visual C + + programming circumstances using C/C + + language.5. We discuss the system's extensibility, and introduce level inspection module. We design the level inspection algorithm process after analyzing the image features. In the end we complete code design in Visual C++ programming circumstances using C/C++ language in the basis of impurities inspection system.6. We use existing conditions to inspect the algorithm function. It proves that the algorithm we proposed is reasonable, and the algorithm can meet desired targets...