Research On Detection Of Conveyor Belt Tearing And Deviation Based On Image Processing

Modern science and technology are changing with each passing day,and the secondary industry is constantly developing towards automation and intelligence.The emergence of conveyor belts just meets the development trend of modern production intelligence.The conveyor belt not only saves labor costs,improves production efficiency,but also regulates the production management of the enterprise.For coal mine production,conveyor belt is an essential transportation force.However,due to the complexity of the working environment of the mine and the heavy workload,the tearing,slipping and deviation of the conveyor belt are the most common failures.Such failures will bring serious economic losses and even casualties to the factory.Therefore,after image preprocessing of mine conveyor belt,this thesis studies the tearing and deviation detection of mine conveyor belts.The detection algorithm of belt tearing and running deviation is studied in detail.The following contents are mainly studied.(1)The conveyor belt image is preprocessed.The image is required by dividing video slice of the mine conveyor belt collected by camera into frame,before the image enhancement processing is performed.Image preprocessing uses a weighted guided filtering algorithm of Two-Dimensional Variational Mode Decomposition(2D-VMD)for image enhancement.By using the advantage of suppressing high frequency noise in the high frequency region,the weighted edge of the low frequency region after decomposition is filtered,and the image enhancement effect is finally achieved,for the purpose of reducing noise,brightening chroma and enhancing detail textures.(2)Aiming at the crack problem of conveyor belt,an edge detection algorithm based on wavelet fusion is proposed.According to the multi-scale decomposition characteristics of wavelet,the image is divided into high frequency and low frequency regions.Multi-scale and multi-direction structural element morphology is used to deal with the low frequency region and the improved Canny operator processes the high-frequency region in order to extract the crack texture information respectively.Finally,the original image is reconstructed by image fusion technique-inverse wavelet.In addition,the image is evaluated and analyzed by parameters such as root mean square error and peak signal-to-noise ratio.The experimental results show that the algorithm is superior to the traditional detection method in noise immunity and crack extraction information.(3)Aiming at the belt off-tracking fault,the algorithm of dynamic programming edge segment detection offset fault is adopted.After the image cumulative Probability Hough Transform(PPHT),the detected lines are positioned by orthogonal decomposition of pixels.The ordering is modeled as Markov chain,the linear ordering is done by dynamic programming.The expected quantity is calculated by the trailing edge probability method,and the straight line detection on both sides of the conveyor belt is completed.Finally,the existence of belt deviation phenomenon is judged according to the straight slope range.(4)MFC(Microsoft basic class library)human-computer interaction interface is built based on the above research content.MFC not only can package the tear and deviation algorithm program,provide a friendly human-computer interaction interface platform,but also set experimental parameters of common image processing.The detection efficiency and precision of conveyor belt are finally improved.