In order to solve the problem of pavement crack automatic detection

a pavement crack image detection algorithm based on cuckoo search is proposed. Firstly

the RGB color image is converted to the HSV space

and the V component is de-noised with total variation model. Then the initial clustering center of the image is calculated by histogram. The algorithm is used as the search domain constraint for iteration of the cuckoo algorithm. Finally

the new clustering center is calculated by the K-means algorithm

and the position of the bird’s nest is updated. The optimal threshold of image segmentation is obtained. The experimental results show that the proposed algorithm has strong denoising ability

good segmentation effect

and can extract pavement cracks well

and its performance is prior to Otsu algorithm.