Deblurring Crack Images Generated by UAV Camera Shake

With the development of computer technology, digital image processing technologies have been applied to many areas of real life, blurred image restoration also has a rapid development, which gives a certain basis and conditions in the bridge crack detection. Aiming at the blurred crack image generated by camera shake, the paper studies the motion blur image restoration algorithms, and explores the parameter estimation methods of motion blur, where the direction and scale of the blur kernel function are estimated from the spectrum of the blurred image. The paper uses different image quality evaluation standards to compare the output, which can choose the best results and gets a more accurate point spread function. This method can obtain clearer crack images and provide more accurate crack information for bridge project research.