Research On Nonuniform Illumination Image Correction And Fault Detection Algorithm For Conveyor Belt

Belt conveyer is a kind of modern production equipment used in continuous transportation, which has the advantages of large capacity, long distance transportation, low energy consumption and so on. It has been widely used in coal, mining,port, metallurgy, chemical,electric power and other fields. The conveyor belt is key components for traction and bearing of Belt conveyer. In the course of using, there are many reasons for the damage to the conveyor belt, including the waste of iron or coal gangue or other obstacles scratch conveyer belt, Improper installation, uneven load because of the position of the material is not positive etc. It will produce wire rope core failure and surface failure, then it will cause a major the safety of the accident, the loss of transport material, equipment be damaged, huge economic losses and casualties, serious impact on safety production. In order to ensure the safe operation of conveyor belt, the research group has developed a nondestructive testing system based on the X-ray of the wire rope core conveyor belt and an on-line monitoring system for conveyor belt fault based on machine vision.In this paper, the algorithm of non-uniform illumination image correction and fault detection is studied for the two systems, we proposed a new kind of non-uniform illumination image illumination X-ray image correction and enhancement algorithm based on Retinex theory. The proposed algorithm firstly implements the low-pass filtering of the X-ray image images by using Butterworth low-pass filter, after denoising and smoothing, the images are obtained; then establish estimated background model of the non-uniform illumination based on Retinex theory; the model is used to get the corrected image; finally, the image contrast is enhanced by using the method of contrast enhancement, and the effect of non-uniform illumination is removed. This paper is to study the conveyor belt damaged image detection algorithm based on machine vision, proposed a new kind of detection algorithm based on machine vision which can be used to realize the image correction and fault detection of conveyor belt. The proposed algorithm firstly implements the low-pass filtering of the acquired images by using Butterworth low-pass filter, then establish estimated background model of the non-uniform illumination based on Retinex theory. Gray scale of the image can be amended evenly. Lastly, by using the combination the machine vision with biological vision and PNCC theory, we can detect the defected area of collected surface image of the conveyor belt. On the MATLAB platform, the program of the above algorithm is written and the experiment is carried out, experimental results have shown that the proposed algorithm which non-uniform illumination X-ray image correction and enhancement based on Retinex theory can effectively correct the non-uniform illumination X image and improve the image contrast, it has the advantages of good correction effect and fast speed; the proposed algorithm which non-uniform illumination image correction and fault detection algorithm based on machine vision can effectively used to correct the uneven gray scale of the surface images and detect the defected area of the surface image. It has been proved that this proposed algorithm has detected fault area clearly.The algorithm we proposed can be applied to the X-ray based wire rope core conveyor belt nondestructive testing system and the on-line monitoring system based on machine vision, realize the correction of non-uniform illumination image illumination image and surface fault detection for conveying belt. To prevent a major the safety of the accident, the loss of transport material, equipment be damaged, huge economic losses and casualties. Improve the safety efficiency of production, it can be widely used in coal, mine. port, metallurgy, chemical, electric power and other fields.