opencv - Emgu.CV GpuSURFDetector against n-images? -
i'm trying extend example below allow matching against more 1 reference image. know how this?
i'm not sure how concat multiple sets of gpumodeldescriptors detector can use. surely there must way compare image against entire set of reference images , not 1 @ time?
the goal load big set of reference images gpu, , compare observed images against references looking (close) match adjustable tolerance/precision. in advance light can shed on this. i've been struggling many months.
example: http://www.emgu.com/wiki/index.php/surf_feature_detector_in_csharp
to knowledge there no out-of-the-box solution task on emgucv.
into detector can use
the role of detector detect key points.
the example takes 2 image, detects key points, produce descriptors each , compares them matching. need modify process of gathering , managing descriptors (for images want use reference) , apply matching of each of observed images. note should store info can image , keypoint each descriptor corresponds to. task not trivial may seems @ first glance.
Comments
Post a Comment