Research On Image Detection Method Of Underwater Dam Cracks

Farmland water conservancy project is a key construction project in China,which is related to the people's livelihood issues in China.As an important project in farmland water conservancy project,the dam project is mainly used to avoid flooding,to withstand floods,and to ensure the safety of people and people along the embankment.However,during the long-term service,the dam is prone to cracks and cause dam body penetration.Difficult to break the embankment.It can be seen that the safety detection of dams is particularly important.The development of machine vision has realized the safe detection of dams by using image processing technology.However,due to the complexity of underwater environment imaging,the image of cracks in underwater dams directly collected will have serious degradation problems.Therefore,image processing technology is used for water.The restoration of the next image becomes an important research topic.Because of the complexity of underwater environment,the image of underwater dam cracks has the problems of blurred edges,high noise and low contrast,which makes it difficult for some classical detection algorithms to accurately detect the target of underwater dam cracks.Therefore,a new underwater image detection algorithm is proposed in this paper.The main research work is as follows:Because of the complexity of underwater environment,the image of underwater dam cracks has the problems of blurred edges,high noise and low contrast,which makes it difficult for some classical detection algorithms to accurately detect the target of underwater dam cracks.Therefore,a new underwater image detection algorithm is proposed in this paper.The main research work is as follows:Firstly,a kind of retinue image enhancement algorithm based on steering filter is proposed to restore the original image.The RGB color space of the underwater dam crack image is transformed into HSV color space.The brightness image is smoothed by steering filter,and the illumination component is estimated.The histogram clipping and Gamma color correction technology are used to calculate the illumination image,and then the Sigmoid color correction technology is used to calculate the illumination image.Finally,the transformation of color space is restored to enhance the image of underwater dam cracks.The algorithm is compared with the typical Retinex algorithm through experiments,and the image quality evaluation criteria of several algorithms are judged.The results show that the Retinex image enhancement algorithm based on steering filter can clearly restore the image of underwater dam cracks,and solve the problem of color bias and "color bias" in the high contrast edge area of Retinex image enhancement.The defect of halo artifact;Secondly,the image of underwater dam crack enhancement is segmented by adaptive maximum variance threshold segmentation method combined with morphological open operation processing method,and the isolated noise points are eliminated by using bwarea open()function threshold setting method;Finally,the clear binary image of underwater dam cracks is extracted from the infinite skeleton,and the linear cracks are refined to a single-pixel width to calculate the required crack parameters.The experimental results show that the method can accurately and effectively detect and extract the crack characteristics of the underwater dam crack image in complex underwater environment.