{"title":"基于改进GMS的航拍图像拼接算法","authors":"K. Yan, Min Han","doi":"10.1109/ICIST.2018.8426189","DOIUrl":null,"url":null,"abstract":"Feature matching is of great importance in the keypoint-based image stitching. Grid-based Motion Statistics (GMS) is a fast and ultra-robust image feature matching algorithm. However the correct matching rate and registration precision of GMS are relatively low. In order to obtain accurate aerial stitching images while ensuring high matching speed, an aerial image mosaic algorithm based on improved GMS is proposed in this paper. Firstly, we apply the ORB algorithm to extract and describe the feature points of the image. Then, GMS-based bidirectional matching is used to acquire the initial matching points. After that, false matches are rejected by constructing epipolar constraint. Finally, we use Random Sample Consensus Algorithm (RANSAC) to calculate the transformation model and fuse the aligning images by weighted average fusion algorithm. Experimental results show that the proposed algorithm has good matching accuracy and registration accuracy while maintaining a low matching time.","PeriodicalId":331555,"journal":{"name":"2018 Eighth International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":"{\"title\":\"Aerial Image Stitching Algorithm Based on Improved GMS\",\"authors\":\"K. Yan, Min Han\",\"doi\":\"10.1109/ICIST.2018.8426189\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Feature matching is of great importance in the keypoint-based image stitching. Grid-based Motion Statistics (GMS) is a fast and ultra-robust image feature matching algorithm. However the correct matching rate and registration precision of GMS are relatively low. In order to obtain accurate aerial stitching images while ensuring high matching speed, an aerial image mosaic algorithm based on improved GMS is proposed in this paper. Firstly, we apply the ORB algorithm to extract and describe the feature points of the image. Then, GMS-based bidirectional matching is used to acquire the initial matching points. After that, false matches are rejected by constructing epipolar constraint. Finally, we use Random Sample Consensus Algorithm (RANSAC) to calculate the transformation model and fuse the aligning images by weighted average fusion algorithm. Experimental results show that the proposed algorithm has good matching accuracy and registration accuracy while maintaining a low matching time.\",\"PeriodicalId\":331555,\"journal\":{\"name\":\"2018 Eighth International Conference on Information Science and Technology (ICIST)\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2018-06-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"6\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2018 Eighth International Conference on Information Science and Technology (ICIST)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICIST.2018.8426189\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 Eighth International Conference on Information Science and Technology (ICIST)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICIST.2018.8426189","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Aerial Image Stitching Algorithm Based on Improved GMS
Feature matching is of great importance in the keypoint-based image stitching. Grid-based Motion Statistics (GMS) is a fast and ultra-robust image feature matching algorithm. However the correct matching rate and registration precision of GMS are relatively low. In order to obtain accurate aerial stitching images while ensuring high matching speed, an aerial image mosaic algorithm based on improved GMS is proposed in this paper. Firstly, we apply the ORB algorithm to extract and describe the feature points of the image. Then, GMS-based bidirectional matching is used to acquire the initial matching points. After that, false matches are rejected by constructing epipolar constraint. Finally, we use Random Sample Consensus Algorithm (RANSAC) to calculate the transformation model and fuse the aligning images by weighted average fusion algorithm. Experimental results show that the proposed algorithm has good matching accuracy and registration accuracy while maintaining a low matching time.