Edge Detection of Car Body Scratches Based on Wavelet Transform and Morphology
In view of the influence of strong local reflection and external noise interference on the surface of the car body,it is impos-sible to accurately extract the effective information of the car body scratch edge.It proposes an edge detection method that combines wavelet transform and mathematical morphology.First of all,in order to reduce the influence of external noise interference such as light in the process of car body scratch edge extraction,the wavelet modulus maximum edge detection method is used for the scratch source image to eliminate some noise interference and improve the accuracy of edge detection and positioning;Optimize the continu-ity of the scratched edge to avoid the loss of effective edge information under multi-scale wavelet detection,and propose an improved mathematical morphology edge detection algorithm;finally,the detection of wavelet modulus maximum and mathematical mor-phology are two methods.The low-frequency components of the two images are fused according to the weight of the image energy dis-tribution to obtain the fused low-frequency component.The corresponding high-frequency component is compared with the absolute value,and the larger of the two is used as the high-frequency component of the fused image.High frequency and low frequency com-ponents are reconstructed to obtain the final scratch edge.Through experimental analysis,the fusion method can guarantee to ob-tain more complete and clear edge feature information while removing noise,the effect is better than traditional edge detection algo-rithms,and it can provide theoretical and technical support for the subsequent precise location and repair of scratches.
万方数据
维普
