An image segmentation method based on Renyi relative entropy and Gaussian distribution

Background: Image segmentation is a necessary prerequisite for many higher level computer vision tasks, such as object recognition, image understanding, and image retrieval. However, the segmentation problem is inherently ill-posed due to the large number of possible partitionings for any single image. So, image segmentation remains one of the major challenges in image analysis. There are also many patents on these problems. This paper introduces a new image segmentation method to readers, and the effectiveness of the new method is demonstrated by experiments. Method: Image segmentation is an important step to obtain quantitative information in image processing and techniques based on thresholding are popular used in practical application. Gaussian distribution is a parametric statistical model, which is frequently employed to characterize the statistical behavior of a process signal in industry. This paper considers the Gaussian distribution to approximate the histogram distribution of an image. A new histogram thresholding segmentation method is presented based on Renyi relative entropy. Results: The experimental results for non-destructive testing image and other type images demonstrate the success of the proposed method, as compared with the alternative thresholding methods. Conclusion: Based on the measurement of Renyi relative entropy and Gaussian distribution, a new image thresholding approach is presented in this paper. The experimental results on various kinds of image show that the proposed method is valid for the task of image segmentation. © 2017 Bentham Science Publishers.