Research Of Image Recognition Technology And Application Based On Color

In recent years,the synthetic leather has been widely used and the industry has developed rapidly.However,due to the difficulty of automatic control of the scraper in the production line of the synthetic leather,it is difficult to achieve the full closed production.Therefore,this results in a large number of unorganized polluting gas emissions,causes environmental pollution and harm the health of workers.The key of the automatic control of the scraper is to solve the problem of release paper joint detection,but due to the movement of the release paper,the color is not fixed,and the existence of stains and damage and other problems,the traditional method is difficult to accurately and automatically detect the release paper joint.Based on the analysis of the characteristics of the release paper and the release paper joint,this paper study on the detection principle and method of the release paper joint based on visual images and use the edge detection technology of color image to achieve online detection and identification for the release paper joint.This paper mainly research the following points:(1)This paper analyzes the technical requirements and the solution methods for the detection of the release paper joint.And according to the characteristics of the image of the release paper joint and the online inspection requirements,this article propose the color image edge detection program and analyze the feasibility of the program.(2)To the online inspection requirements of the release paper joint,this paper study the adaptability of the commonly used color image edge detection algorithm based on the single component.And this article analyze and simulate the performance of the traditional gradient operator for edge detection.Then for its shortcomings,this paper propose improved algorithm about combined gradient operator and gradient operator with a combination of mathematical morphology.Through experimental analysis,the traditional algorithm and the improved algorithm have a certain universality to the edge detection,but can not meet the requirements for the online detection of the release paper joint.(3)In order to improve the detection accuracy,this paper studies the color image edge detection technology of component comprehensive.Based on the typical RGB and HSV color space,this article analyze and simulate the characteristics of the color image edge detection algorithm of the three component comprehensive.In order to improve the detection effect,the algorithm based on RGB is studied by fusing the algorithm of combined gradient operator and the other one is based on the characteristics of HSV space to realize the combination of gradient operator and color formula.Experiments show that the algorithm testing results are better,but because of the processing time is too long,can not meet the requirements for the online detection of the release paper joint.(4)Based on the study of the algorithm of the single component and the component comprehensive and combined with the color feature of the release paper,this paper specifically analyze the distribution trend of three component color value of RGB and propose a processing algorithm based on RGB.By the processing of three component pixel values,this algorithm can find the edge points of severe mutation and judge the joints from the release paper.Experiments show that the algorithm has high accuracy and fast processing speed,can meet the requirements of the automatic carrying scraper system.And the algorithm has been applied to the actual scene and has good adaptability.