opencv-api FlannBasedMatcher
阿新 • • 發佈：2018-12-22
Fast Library forApproximate Nearest Neighbors
1.建立物件
<FlannBasedMatcher object> = cv.FlannBasedMatcher( [, indexParams[, searchParams]] )
indexParams
| 引數 | 描述 |
|---|---|
| FLANN_INDEX_LINEAR | 線性暴力(brute-force)搜尋 |
| FLANN_INDEX_KDTREE | 隨機kd樹，平行搜尋。預設trees=4 |
| FLANN_INDEX_KMEANS | 層次k均值樹。預設branching=32，iterations=11，centers_init = CENTERS_RANDOM, cb_index =0.2 |
| FLANN_INDEX_COMPOSITE | 隨機kd樹和層次k均值樹來構建索引。預設trees =4，branching =32，iterations =11，centers_init = CENTERS_RANDOM,cb_index =0.2 |
| FLANN_INDEX_KDTREE_SINGLE | |
| FLANN_INDEX_HIERARCHICAL | |
| FLANN_INDEX_LSH | multi-probe LSH方法 |
| FLANN_INDEX_SAVED | |
| FLANN_INDEX_AUTOTUNED | |
| LINEAR | |
| KDTREE | |
| KMEANS | |
| COMPOSITE | |
| KDTREE_SINGLE | |
| SAVED | |
| AUTOTUNED | 自動選取，以提供最好的效能 |
searchParams
SearchParams (int checks=32, float eps=0, bool sorted=true)
| 引數 | 描述 |
|---|---|
| checks | 預設32 |
| eps | 預設為0 |
| sorted | 預設True |
2.匹配方法
FlannBasedMatcher物件繼承了cv::DescriptorMatcher.
matches = cv.DescriptorMatcher.knnMatch( queryDescriptors, trainDescriptors, k[, mask[, compactResult]] ) matches = cv.DescriptorMatcher.knnMatch( queryDescriptors, k[, masks[, compactResult]] )
| 引數 | 描述 |
|---|---|
| queryDescriptors | 原圖 |
| trainDescriptors | 搜尋的圖片 |
| matches | 匹配的結果 |
| k | Count of best matches found per each query descriptor or less if a query descriptor has less than k possible matches in total.設定閾值，越高精度越高，匹配的數量越少 |
| masks | Set of masks. |
| compactResult | Parameter used when the mask (or masks) is not empty. |
matches
| 引數 | 描述 |
|---|---|
| distance | |
| imgIdx | 訓練圖片的索引 |
| queryIdx | query descriptor index |
| trainIdx | train descriptor index |