Research on Segmentation Algorithm of Gray Inhomogeneous Image Based on Cauchy Distribution
Image segmentation has a constructive position in image engineering and other fields. Among them, the research on the segmentation of uneven grayscale images is particularly important. This is due to the fact that uneven grayscale images widely exist in the real world, such as medical images, remote...
Saved in:
Published in | 2020 5th International Conference on Information Science, Computer Technology and Transportation (ISCTT) pp. 287 - 291 |
---|---|
Main Author | |
Format | Conference Proceeding |
Language | English |
Published |
IEEE
01.11.2020
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Image segmentation has a constructive position in image engineering and other fields. Among them, the research on the segmentation of uneven grayscale images is particularly important. This is due to the fact that uneven grayscale images widely exist in the real world, such as medical images, remote sensing images, and video surveillance. However, the traditional image segmentation algorithm ignores the unevenness of the gray level of the image, and the effect of such image segmentation is poor. Therefore, this paper proposes a gray-scale uneven image segmentation algorithm based on Cauchy distribution. Based on the RSF (region-scalable fitting) active contour model, this algorithm creates a new kernel function based on the Cauchy distribution, which is the absolute value of the difference between the two Cauchy distributions. On this basis, the energy functional is re-established to fit the gray value of the image inside and outside the contour, and the contour penalty item is added. Finally, the level set theory is used to convert the energy functional into a level set form and add a level set regularization term, and use the gradient descent method to minimize the energy functional. The experimental results show that using the method in this paper to segment the gray-scale uneven image has higher segmentation accuracy and segmentation efficiency, and the segmentation speed is increased by nearly 50%. |
---|---|
DOI: | 10.1109/ISCTT51595.2020.00057 |