Multi-exposure images matching based on improved sift

作者：Zhang Ming; Hu Jiehua; Zheng Caixia^*; Wang Jianzhong

来源：ICIC Express Letters, Part B: Applications , 2011, 2(5): 1175-1180.

• Brightness-space images
• Contrast context histogram
• Contrast stretching
• Descriptors
• Keypoints
• Local region
• Multi-exposure images
• Non-uniform illumination
• Scale invariant feature transforms
• Sift

This paper presents an efficient method for multi-exposure images matching. Firstly, in the purpose of obtaining more consistent feature points under large nonuniform illumination changes, we use scale invariant feature transform (SIFT) to extract keypoints from the brightness-space images which are built by the contrast stretching function with various parameters. Then, to overcome the disadvantages of SIFT descriptor lacking contrast information, contrast context histogram (CCH) is incorporated into the SIFT descriptor. CCH is the distributions of the contrast within a local region and it is more discriminating for multi-expo sure image matching. The experimental results demonstrate that the proposed method improves the SIFT performance for multi-expo sure images matching significantly.

• 出版日期2011
• 单位东北师范大学