| The dam project occupies an important position in the construction of water conservancy projects in my country,and is a key construction project of water conservancy projects in my country.Since the cracks of dams in dyke engineering are common safety hazards,the detection of cracks in underwater dams is an important part of the safety precautions of dam engineering.The development of computer vision and image processing technology has promoted the improvement of crack detection technology.Underwater image acquisition equipment is used to detect cracks in underwater dams.However,due to the scattering effect and attenuation characteristics of light in the underwater scene,water image of the bottom crack usually has problems such as low-illuminance blur and color shift,which is not conducive to the detection of underwater dam crack images.Therefore,it is an important research to accurately and effectively extract the crack information of underwater dams.There are many problems in underwater dam crack image,such as low illumination,blurred edge details and color deviation.The traditional image detection algorithm can not accurately and effectively extract the crack information in the underwater dam crack image,according to the principle of underwater optical imaging and the problems of underwater image,this paper proposes a new image detection algorithm for low illumination underwater dam cracks.The basic idea of algorithm research is as follows.Firstly,according to the characteristics of underwater dam crack image and the degradation problem of the underwater dam crack image,the MSR algorithm is improved in a targeted manner.The improved MSR algorithm with adaptive weight selection is divided by the average gradient of the pixels in the image.There are seven threshold ranges,each pixel within the average gradient range is enhanced by the MSR algorithm with the corresponding preset scale weight factor,and then the brightness of the image is adjusted through Gamma correction to obtain an enhanced underwater crack image.Secondly,perform image segmentation and crack extraction on the enhanced underwater dam crack image,and analyze the low-illuminance blur in the underwater dam crack image in advance.The existing crack detection methods cannot accurately and completely extract the underwater The problem of thin linear cracks and wide linear cracks existing in dams,combined with FCM clustering algorithm,edge detection method,and improved maximum between-class variance(Otsu)algorithm to perform image segmentation and binarization of the crack image,effectively suppress The noise interference in the image is taken into account,the complete small and wide linear dam crack information is accurately extracted,and the parameter value of the dam crack information is estimated.Finally,the experimental results show that the proposed image detection method can accurately and effectively extract the target information of fine dam cracks and wide linear dam cracks,and estimate the length,width,area and tilt angle of dam cracks. |
