Research on Image Matching Evaluation Algorithm for Visual SLAM
| Autor: | Weishan Liang, Yun Zhao, Yibo Wang |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Similarity (geometry)
Pixel Computer science business.industry Binary image ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Inter frame Simultaneous localization and mapping Image (mathematics) Hausdorff distance Feature (computer vision) Computer Science::Computer Vision and Pattern Recognition Computer vision Artificial intelligence business |
| Zdroj: | 2021 International Conference on Electronic Information Engineering and Computer Science (EIECS). |
| DOI: | 10.1109/eiecs53707.2021.9588110 |
| Popis: | VO algorithm is the key to the application of vision slam. In view of the complexity and time-consuming of VO feature point method, the consistency of image matching is studied theoretically, and the image consistency based on pixel, content and feature is defined respectively. On this basis, the potential problems of pixel based image matching algorithm are analyzed, and a feature-based image matching algorithm is proposed. This algorithm makes full use of the advantage of Hausdorff distance to solve the similarity of binary image, and achieves a relatively reliable evaluation of image matching, which provides a basis for the reconstruction of image inter frame motion. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|