The Algorithm Of Full Box Detection Based On Machine Vision And The System Development

The specific target in the image of real-time online testing is the key technologies in machine vision applications. Specific target detection by judging whether the target and background in determining the location of a target is composed of two parts, Target model and the design of the detection algorithm will be directly related with the efficiency, accuracy and stability of the test results. Therefore, Studying on real-time, precise detection algorithm of target has been the research hotspot in the field of machine vision and is also the research difficulties. Target detection algorithm is essentially a kind of seeking optimal solution. The existing mature target detection algorithm based on adding various heuristic optimization algorithm has received more and more attention.Hough transforming is mainly good at finding the straight line in the image, but it is not obvious to find round effect on the image and find the optimal solution for too long. The accuracy is low.Election Campaign Algorithm (ECA) was a new heuristic optimization algorithm, which simulated the pursuit of highest support rate in election activities. ECA was almost used in engineering fields.In this paper, it is based on Hough transform and the election algorithm respective characteristic that take on the Hough transform which is improved and discover the new method of detection image circle. The main content is as follows:First of all, introducing the election algorithm thought, framework and features, An overview of the election algorithm calculation process. Genetic algorithm is used in the standard test functions of election algorithm validation.Secondly, the value of the bench mark functions optimization experiments analyze the relationship between the campaign algorithm parameters and the performance of the algorithm.By the number of each campaign to change the election algorithm to study the impact of the performance of the algorithm, provide the basis for the practical application of the election algorithm parameter selection. Based on standard test function numerical optimization experimental results show that the change of control parameters, can improve campaign performance of the algorithm can improve the performance and the search efficiency of algorithm optimization.Thirdly, an overview of the Hough transform definition, Hough transform for detecting straight line and circle detection method, which based on the election algorithm Hough transform, and shot on the algorithm performance and effect of experiment and analysis.Finally, in the OpenCV image processing functions on the basis of the based on the election algorithm Hough transform algorithm based on combination of Microsoft, Visual Studio2005platform to build a machine vision system, realizes the real-time target detection, and verified. Test the performance and effectiveness of the algorithm to use the site to take pictures.