Movement Invariants-based Algorithm for Medical Image Tilt Correction

摘要

In this paper, the edge detection for a medical image is performed based on Sobel operator, and the bounding box is obtained, by which the effective medical sub-image is extracted. Then, the centroid and the normalized central moments of the medical sub-image are calculated, and the rotation angle α is obtained by minimizing the second-order central moment based on its rotation invariance. Finally, the whole medical image is rotated around the centroid by —α to correct the tilted image. Furthermore, inspired by the uniformity degree of the image, the rotation angle α is revised, which achieves a better correction effect and performance. The experimental results show that the proposed algorithms are fairly reliable and accurate for the determination of tilt angles, and are practical and effective tilt correction techniques.