Local reverse entropy weighted LBF model solving by Split Bregman for image segmentation
- PDF / 6,344,607 Bytes
- 25 Pages / 439.37 x 666.142 pts Page_size
- 77 Downloads / 185 Views
Local reverse entropy weighted LBF model solving by Split Bregman for image segmentation Dengwei Wang 1,2 Received: 8 April 2019 / Revised: 29 March 2020 / Accepted: 22 May 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
In this paper, an efficient level set model is proposed for image segmentation. Firstly, the original local binary fitting (LBF) model is redefined as a weighted energy integral, whose weight coefficient is the fast local reverse entropy of the image, and the total energy functional is then incorporated into a variational level set formulation. Secondly, the global convex segmentation method is used to construct a simplified convex segmentation model, at the same time, the edge information obtained by an edge indicator function is embedded into the total variation norm to further enhance the model’s target capture capability. Thirdly, the Split Bregman method is introduced to solve the generated convex optimization problem. Experimental results on synthetic and real images demonstrate that the proposed model has considerable improvements in terms of quantitative evaluation (being verified on the complete PASCAL VOC 2012 dataset), convergence rate, sensitivity to initial contour and robustness to noise interference compared with the state-of-the-art models. We also compare the proposed model with the famous FCN and Mask R-CNN, and make a special analysis on the adaptability of our method to occluded targets. Keywords Local binary fitting . Level set . Local reverse entropy . Split Bregman . Segmentation
1 Introduction Image segmentation is an important intermediate step in the field of computer vison, its output quality is directly related to the accuracy and effectiveness of tasks such as target detection and recognition. Among various types of image segmentation technologies, the geometric active
* Dengwei Wang [email protected]
1
School of Aeronautics and Astronautics, University of Electronic Science and Technology of China, Chengdu 611731, China
2
Aircraft Swarm Intelligent Sensing and Cooperative Control Key Laboratory of Sichuan Province, Chengdu 611731, China
Multimedia Tools and Applications
contour models (the models that introduce the idea of level set formulation) have been widely used due to their ability to handle topology changes and output closed target contours. We can roughly divide the existing segmentation methods based on geometric active contour models into two categories: edge-based methods [4, 13, 18, 21] and region-based methods [1, 5, 8, 11, 16, 20]. The external energy terms (corresponding to the driving force of the active contour) of the edge-based methods are based on the edge gradient information of the image. For example, by associating the geodesic distance with the active contour and using the image gradients as the core power source, Caselles et al. [4] proposed the classical geodesic active contour (GAC) model, which has achieved great success in the segmentation applications of high-quality image data. While the extern
Data Loading...
