Research on fault detection of bucket teeth of mining excavator

This paper studies the bucket tooth fault detection of mining excavator bucket, combines digital image processing, computer vision and other technologies, integrates the target detection technology and image segmentation technology based on deep learning into the bucket tooth fault detection of mining excavator, and designs a complete set of bucket tooth fault detection system of mining excavator to realize the fault detection and fault alarm function of bucket tooth of mining excavator. The main research contents are as follows:

(1) According to the working conditions and harsh working environment of the excavator, the hardware system and software system of the visual detection device of the bucket are designed. Combined with and analyzing the working characteristics of the shovel, the installation and implementation of the detection system are designed. Finally, according to the method used in this paper, a certain number of data sets are established to complete the annotation of data and the processing of data sets.

(2) The object detection method based on deep learning is used to detect the bucket teeth of the bucket, obtain the position information of the bucket teeth of the mining excavator in the image, and realize the preliminary judgment of fault detection. This part mainly uses the fast r-cnn target detection framework and adjusts it in combination with the characteristics of bucket tooth target detection of mining excavator, so that it can accurately detect the bucket teeth of mining excavator bucket, so as to achieve the purpose of preliminary detection of bucket tooth fault.

(3) After obtaining the position information of the bucket teeth of the mining excavator in the image through target detection, the full convolution neural network is used to segment the image of the bucket teeth of the mining excavator, the contour of the output label image is extracted, the standardized Fourier descriptor is used as the contour feature, and the Euclidean distance is used to distinguish the bucket teeth of the end mining excavator from the middle bucket teeth, combined with the position relationship between the bucket teeth of the mining excavator, Determine the position of the bucket teeth of the faulty mining excavator, calculate the length of the bucket teeth of the faulty mining excavator by using simple image processing, compare the bucket teeth of the normal mining excavator, and judge the wear degree of the bucket teeth of the mining excavator.

(4) According to the requirements of the bucket tooth detection system of mining excavator, the function of the system is designed. Combined with the user interface design of PyQt4 of python, the visual interface of the bucket tooth detection system of mining excavator bucket is designed to realize the functions of image reading, bucket tooth fault detection of mining excavator, fault alarm, result visualization and so on.

spacer