Research And Implementation Of Bamboo Defect Detection Based On LabVIEW

With the development of digital image processing technology, the detection technology of computer vision has become a hot issue today. The study applies to theory of digital image processing of bamboo image defect detection and analyzes four kinds of common defect detection algorithms. Based on the algorithms, a new method of bamboo image defect detection is put forward using the mixing programming of Lab VIEW and Matlab.Firstly, this paper uses the scanning method and the minimum circumscribed rectangle method to rotate, translate and match the target position of the tilt image, then establishes a cartesian coordinate system to further complete the digital image processing with coloring image grayscale, image binarization, and feature extraction and so on. With the above pretreatment, it adopts contour detection to check contour defects, employs the algorithm of image subtraction method and gray correlation detection to detect the bamboo image texture defects, and further uses the gray value statistics method to detect the precision of lamination defect. To judge the eligibility of the bamboo, the grey value of bamboo in the target area is extracted precisely, and threshold and standard deviation of grey value is analyzed and compared through Matlab. The above several methods are combined to detect texture and contour defects. The above several methods are combined to test outline and bamboo defects and so on.Then relative experiments have been carried on the detection of contour defect and texture defect, especially algorithm of texture defect based on the analysis of gray value statistics. The algorithm not only is carried on the Lab VIEW platform, but also is carried out in conjunction with Matlab. In addition, this mixing programming overcomes some difficulties of the uneven illumination and positioning inaccuracy, and has more accuracy on target localization and gray value statistics. Combined with program modules of the bamboo contour defects, texture and statistical method of gray value spot defects, the detections of bamboo defects are achieved in this paper.Finally, on the LabVIEW, bamboo defect detection is designed. The samples of1000bamboo images are collected and filtered by artificial selection and program selection for defect of bamboo respectively. The experimental results shown that unbound Matlab programming defect detection rate comes to98.65%, while combined with Matlab mixed programming system, the detection rate increases by0.45%and reaches to99.10%. According to the experimental results, it can show that the use of hybrid programming is more practical and feasible, laying the foundation for industrial automation.