https://github.com/seravee08/WideBaselineFeatureMatcher_PAMI
Window10 CUDA7.0 GTX1060
测试成功！
Microsoft Windows [版本 10.0.17134.345] (c) 2018 Microsoft Corporation。保留所有权利。 E:\yangninghua\paper-works\CODE-GMS\WideBaselineFeatureMatcher_PAMI-master\CODE>code ========================= CODE Useage Instructions: ========================= "-mach" takes in 2 sift files, output a .mat binary matching "-blat" takes in 2 sift files and automatically search for the .mat file, output final matchings in the first sift files's folder "-sift" takes in a path to image or to a .txt file, output the sift features in the same folder as the pictures "-pwis" takes in a path to a .txt file, computes all to all matches and output the final matchings in the same folder as the .txt file "-full" takes in 2 paths to 2 .txt files, computes matches between lists, gist can be enabled "-gist" enables gist mode, uses gist to pre-filter the candidate images "-srdy" indicates you already have sift files for the images "-grdy" indicates you already have gist files for the images "-umst" indicates whehter you want to use minimum spanning tree "-iter" the number of iterations for minimum spanning tree "-nbor" the number of neighbors to match for minimum spanning tree "-disp" displays the matchings visually "-stpt" used in pwise mode, indicates which image to continue "-pcom" computes sift and gist features for all images in the given list, output as two .dat files in the current folder (not recommanded) "-indb" indicates the first input path "-inqy" indicates the second input path "-nmal" output normal matching coordinates "-rept" turn on full timing report Example usage: C:/CODE -sift -indb C:/image.txt E:\yangninghua\paper-works\CODE-GMS\WideBaselineFeatureMatcher_PAMI-master\CODE>CODE.exe -sift -indb E:/yangninghua/paper-works/CODE-GMS/WideBaselineFeatureMatcher_PAMI-master/CODE/image_list.txt Using best device : device 0 ... ======================================== CUDA device name: GeForce GTX 1060 Compute capability: 6.1 Global memory is: 6144MB Warning: no settings for this device found, please contact the author ... Engaging (CUDA_ASIFT) ... Extracting Asift ... GPU Asift time is: 87.5662 second(s) 75790 ASIFT keypoints are detected. GPU Asift time is: 1.08463 second(s) 74174 ASIFT keypoints are detected. GPU Asift time is: 1.08285 second(s) 75546 ASIFT keypoints are detected. GPU Asift time is: 1.01161 second(s) 74103 ASIFT keypoints are detected. GPU Asift time is: 1.03704 second(s) 77856 ASIFT keypoints are detected. Time elapsed is: 94.502 s E:\yangninghua\paper-works\CODE-GMS\WideBaselineFeatureMatcher_PAMI-master\CODE>CODE -mach -indb E:/yangninghua/paper-works/CODE-GMS/WideBaselineFeatureMatcher_PAMI-master/CODE/image/B21.sift -inqy E:/yangninghua/paper-works/CODE-GMS/WideBaselineFeatureMatcher_PAMI-master/CODE/image/B22.sift Number of matches: 75790 Matching Time is: 2.204 s Time elapsed is: 2.328 s E:\yangninghua\paper-works\CODE-GMS\WideBaselineFeatureMatcher_PAMI-master\CODE>CODE -mach -blat -indb E:/yangninghua/paper-works/CODE-GMS/WideBaselineFeatureMatcher_PAMI-master/CODE/image/B21.sift -inqy E:/yangninghua/paper-works/CODE-GMS/WideBaselineFeatureMatcher_PAMI-master/CODE/image/B22.sift Number of matches: 75790 Matching Time is: 2.265 s Warning: validated number of matches : 15134 Time elapsed is: 4.584 s E:\yangninghua\paper-works\CODE-GMS\WideBaselineFeatureMatcher_PAMI-master\CODE>
