Name
Abstract | ||
|---|---|---|
This A novel level set method based on prior shape is proposed in view of the problem which occurs when the existing level set method segments the images with complicated background or intensity inhomogeneity. The method employs Deep Boltzmann Machine to learn prior shape which can satisfy the global deformation and local deformation. Then by use of the variational level set method and the local Gaussian distribution fitting energy, the image energy term is represented by local mean and local variance of image and the prior shape energy is introduced to construct the final energy term of image segmentation model. Experimental results show that our model can segment images under different prior shapes, intensity inhomogeneity and partial occlusion and has the advantages in terms of computational accuracy and efficiency. |
Year | DOI | Venue |
|---|---|---|
2017 | 10.1109/CISP-BMEI.2017.8301981 | 2017 10th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI) |
Keywords | Field | DocType |
Face Segmentation,Deep Learning,Level Set,Prior Shape | Boltzmann machine,Pattern recognition,Level set method,Computer science,Segmentation,Level set,Image segmentation,Distribution fitting,Gaussian,Artificial intelligence,Deep learning | Conference |
ISBN | Citations | PageRank |
978-1-5386-1938-4 | 0 | 0.34 |
References | Authors | |
4 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Xiaoling Wu | 1 | 54 | 10.26 |
| Ji Zhao | 2 | 11 | 3.91 |
| Huibin Wang | 3 | 29 | 10.99 |