Recognition Algorithm Research And Software Development Of Online Statistical Image For Enamel Production Line

Image recognition technology is one of the key technologies to realize the intelligent manufacturing in the future, which also has a wide application prospect for the transformation of traditional production line. The research object of this paper is the online statistical system of an enamel product line, and the techniques of image preprocessing, feature detection and image recognition related to the system are studied. Finally, using the GUI toolkit of Matlab realize a simple online statistical prototype system. The work of this paper is mainly from the following aspects:Firstly, the image preprocessing algorithm is studied. Through contrast and analysis of the treatment effect of histogram equalization, mean filtering, piecewise linear transform, gamma transform and other image enhancement algorithm,determine that the piecewise linear transformation algorithm is used to process the image of single product. For the classification and counting of the products in the production of mixed products, the method of color extraction based on RGB color space and HSV color space is studied. By comparing the extraction effects of the two methods, the method of color extraction based on HSV color space is selected to extract the target region in the mixed product image.Secondly, the method for detecting the characteristics of enamel products is studied. For collected single product image, this paper presents a fast ellipse detection method, which realizes the rapid identification of product features. Aiming at the problem of image recognition in the production of mixed products, the detection of the position and size of the target region, segmentation of connected regions and perspective transform are studied, which realize the image recognition and product classification statistics under the condition of mixed product production.Finally, the design of hardware and software system is completed. And the online statistical image recognition prototype system is developed based on Matlab,which can accurately and quickly complete the statistics of enamel products on the production line.