A new edge detection method of magnetic flux leakage image based on wavelet multi-scale registration and modulus maximum
Weichen Xu, Xi Liu, Zhen Dai · 2018
In this paper, a new edge detection method based on wavelet multi-scale registration and modulus maximum is proposed to better detect the edge of the defect magnetic flux image. Firstly, the axial image A and the radial image B are respectively carried out with binary wavelet transform in the selected scale. The modulus and phase angles of each image point (x, y) are then calculated and combined into a new magnitude to represent the correlation of the image points (x, y). A new universal edge correlation criterion is defined as a measure of similarity in the registration process. Finally, wavelet edge and modulus maximum methods are used to register and fuse edges of images to reconstruct edges. The experimental results show the effectiveness of the proposed method.