Research On Crack Detection System Of Concrete Prefabricated Component Based On Image Processing

As the assembly of the building with a fast construction,stable and reliable quality,energy saving and environmental protection,it has been rapid development in our country.Concrete prefabricated components is an important part of fabricated building.When the concrete prefabricated parts of the width of more than 0.2 cracks will seriously affect the quality of the assembly of the building.At present,The artificial method is mainly used to detect the cracks.it's not only low efficiency but also low precision.In order to improve the current situation,studied a method of crack detection of concrete prefabricated components based on digital image processing,this method can effectively detect the cracks.For the crack image in the collection process easily affected by the external environmental caused by the weak contrast,image blur and other defects.At the same time image segmentation algorithm has an important effect on the crack detection accuracy,so the segmentation algorithm of crack image segmentation is mainly studied.Firstly,using grayscale,grayscale enhancement,and adaptive median filtering to preprocess the crack image to reduce the influence of the above defects.Secondly using pulse coupled neural network(PCNN)model to simulate the crack image segmentation.Aiming at the instability of PCNN model,the PCNN algorithm is improved by genetic algorithm and minimum error criterion.And the identification of cracks in the image is realized by combining the geometrical features of the fracture.And calculate the length and average width of the identified crack.According to the appeals fracture detection methods,design a set of concrete crack prefabricated detection system,The feasibility and validity of the system are verified by experiments.