Research On The Classification And Recognition Technology Of Waste Carton Based On Machine Vision

The recycling and reuse of waste cartons not only alleviates the shortage of papermaking raw materials to a great extent,but also alleviates the environmental pollution and has a broad application prospect.However,many of the cardboard boxes exist coated,tape wrapping phenomenon,in the field of express delivery,the packaging and transportation of goods will use a lot of tape to carton reinforcement and protection,in the production field,most of the products will use coated cardboard boxes in order to moistureproof,anti-pollution and reinforcement.When this kind of waste carton enters into the paper factory’s reproduction process,its plastics and adhesives will affect the purity of secondary fiber,and even cause equipment damage,so it is necessary to classify the waste carton,remove the film-coated waste carton,and separately treat it,and then use it for paper pulp.At present,laminating waste carton sorting mainly rely on the human eye for identification,due to the human subjective factors and the inevitable fatigue of human eyes,waste carton classification efficiency is low and error detection rate is high,which will lead to secondary fiber purity greatly reduced,affecting the quality of paper.Machine vision technology has the advantages of high efficiency,automation and intelligence,which can replace human eye detection and meet the requirements of efficient classification and identification of coated waste cartons.In this paper,based on machine vision technology,a kind of classification and identification technology of paper-coated waste carton is studied.Aiming at the problem that the existing image enhancement algorithm is difficult to solve this collection of uneven illumination image,an improved MSRCR color image enhancement algorithm is proposed for image enhancement,followed by feature extraction for the waste carton,and finally support vector machine(SVM)classifier is used for classification and recognition of the waste carton.The main work of this paper is as follows:1.According to the actual identification needs and the analysis of the types of waste cartons,the selection of each component of the image collection system(CCD industrial camera,lens,light source and INDUSTRIAL computer)is conducted,and a reasonable image collection platform is built to collect a large number of images of different types of waste cartons.2.In order to reduce the workload of image processing and improve the efficiency of program operation,equal interval sampling technology is adopted to compress images.The Retinex theory based image enhancement algorithm and the improved MSRCR color image enhancement algorithm were used to enhance the image respectively,and the effectiveness of the improved algorithm was verified.Gaussian filter is used to reduce image noise.The segmentation method and morphological processing of the image provide accurate basic support for the next step of feature extraction.3.For the classification and recognition of waste cartons,extract the feature values of processed images by using texture features and morphological features.Support vector machine(SVM)algorithm was used to identify and classify waste cartons.The experimental results show that this method can effectively classify the images with high accuracy.