work on feature_homography.py: multiple targets

samples/python2/common.py

@@ -6,6 +6,12 @@ import itertools as it
 
 image_extensions = ['.bmp', '.jpg', '.jpeg', '.png', '.tif', '.tiff', '.pbm', '.pgm', '.ppm']
 
+class Bunch(object):
+    def __init__(self, **kw):
+        self.__dict__.update(kw)
+    def __str__(self):
+        return str(self.__dict__)
+
 def splitfn(fn):
     path, fn = os.path.split(fn)
     name, ext = os.path.splitext(fn)

samples/python2/feature_homography.py

@@ -20,7 +20,7 @@ import cv2
 import video
 import common
 from collections import namedtuple
-from common import getsize
+from common import getsize, Bunch
 
     
 FLANN_INDEX_KDTREE = 1
@@ -35,11 +35,11 @@ MIN_MATCH_COUNT = 10
 
 ar_verts = np.float32([[0, 0, 0], [0, 1, 0], [1, 1, 0], [1, 0, 0],
                        [0, 0, 1], [0, 1, 1], [1, 1, 1], [1, 0, 1], 
-                       [0.5, 0.5, 2]])
+                       [0, 0.5, 2], [1, 0.5, 2]])
 ar_edges = [(0, 1), (1, 2), (2, 3), (3, 0), 
             (4, 5), (5, 6), (6, 7), (7, 4),
             (0, 4), (1, 5), (2, 6), (3, 7), 
-            (4, 8), (5, 8), (6, 8), (7, 8)]
+            (4, 8), (5, 8), (6, 9), (7, 9), (8, 9)]
 
 
 
@@ -53,7 +53,7 @@ class App:
         self.cap = video.create_capture(src)
         self.frame = None
         self.paused = False
-        self.ref_frame  = None
+        self.ref_frames = []
 
         self.detector = cv2.ORB( nfeatures = 1000 )
         self.matcher = cv2.FlannBasedMatcher(flann_params, {})  # bug : need to pass empty dict (#1329)
@@ -66,13 +66,19 @@ class App:
         if len(self.frame_desc) < MIN_MATCH_COUNT or len(self.frame_desc) < MIN_MATCH_COUNT:
             return
         
-        raw_matches = self.matcher.knnMatch(self.frame_desc, k = 2)
-        p0, p1 = [], []
-        for m in raw_matches:
-            if len(m) == 2 and m[0].distance < m[1].distance * 0.75:
-                m = m[0]
-                p0.append( self.ref_points[m.trainIdx].pt )  # queryIdx
-                p1.append( self.frame_points[m.queryIdx].pt )
+        matches = self.matcher.knnMatch(self.frame_desc, k = 2)
+        matches = [m[0] for m in matches if len(m) == 2 and m[0].distance < m[1].distance * 0.75]
+        if len(matches) < MIN_MATCH_COUNT:
+            return
+        img_ids = [m.imgIdx for m in matches]
+        match_counts = np.bincount(img_ids, minlength=len(self.ref_frames))
+        bast_id = match_counts.argmax()
+        if match_counts[bast_id] < MIN_MATCH_COUNT:
+            return
+        ref_frame = self.ref_frames[bast_id]
+        matches = [m for m in matches if m.imgIdx == bast_id]
+        p0 = [ref_frame.points[m.trainIdx].pt for m in matches]
+        p1 = [self.frame_points[m.queryIdx].pt for m in matches]
         p0, p1 = np.float32((p0, p1))
         if len(p0) < MIN_MATCH_COUNT:
             return
@@ -82,22 +88,28 @@ class App:
         if status.sum() < MIN_MATCH_COUNT:
             return
         p0, p1 = p0[status], p1[status]
-        return p0, p1, H
+        return ref_frame, p0, p1, H
 
 
     def on_frame(self, vis):
         match = self.match_frames()
         if match is None:
             return
+        
         w, h = getsize(self.frame)
-        p0, p1, H = match
-        for (x0, y0), (x1, y1) in zip(np.int32(p0), np.int32(p1)):
-            cv2.line(vis, (x0+w, y0), (x1, y1), (0, 255, 0))
-        x0, y0, x1, y1 = self.ref_rect
+        ref_frame, p0, p1, H = match
+        vis[:h,w:] = ref_frame.frame
+        draw_keypoints(vis[:,w:], ref_frame.points)
+        x0, y0, x1, y1 = ref_frame.rect
+        cv2.rectangle(vis, (x0+w, y0), (x1+w, y1), (0, 255, 0), 2)
         corners0 = np.float32([[x0, y0], [x1, y0], [x1, y1], [x0, y1]])
         img_corners = cv2.perspectiveTransform(corners0.reshape(1, -1, 2), H)
         cv2.polylines(vis, [np.int32(img_corners)], True, (255, 255, 255), 2)
 
+        for (x0, y0), (x1, y1) in zip(np.int32(p0), np.int32(p1)):
+            cv2.line(vis, (x0+w, y0), (x1, y1), (0, 255, 0))
+
+        '''
         corners3d = np.hstack([corners0, np.zeros((4, 1), np.float32)])
         fx = 0.9
         K = np.float64([[fx*w, 0, 0.5*(w-1)],
@@ -110,21 +122,19 @@ class App:
         for i, j in ar_edges:
             (x0, y0), (x1, y1) = verts[i], verts[j]
             cv2.line(vis, (int(x0), int(y0)), (int(x1), int(y1)), (255, 255, 0), 2)
-
+        '''
     def on_rect(self, rect):
         x0, y0, x1, y1 = rect
-        self.ref_frame = self.frame.copy()
-        self.ref_rect = rect
         points, descs = [], []
         for kp, desc in zip(self.frame_points, self.frame_desc):
             x, y = kp.pt
             if x0 <= x <= x1 and y0 <= y <= y1:
                 points.append(kp)
                 descs.append(desc)
-        self.ref_points, self.ref_descs = points, np.uint8(descs)
-
-        self.matcher.clear()
-        self.matcher.add([self.ref_descs])
+        descs = np.uint8(descs)
+        frame_data = Bunch(frame = self.frame, rect=rect, points = points, descs=descs)
+        self.ref_frames.append(frame_data)
+        self.matcher.add([descs])
 
     def run(self):
         while True:
@@ -141,14 +151,9 @@ class App:
             w, h = getsize(self.frame)
             vis = np.zeros((h, w*2, 3), np.uint8)
             vis[:h,:w] = self.frame
-            if self.ref_frame is not None:
-                vis[:h,w:] = self.ref_frame
-                x0, y0, x1, y1 = self.ref_rect
-                cv2.rectangle(vis, (x0+w, y0), (x1+w, y1), (0, 255, 0), 2)
-                draw_keypoints(vis[:,w:], self.ref_points)
             draw_keypoints(vis, self.frame_points)
 
-            if playing and self.ref_frame is not None:
+            if playing:
                 self.on_frame(vis)
             
             self.rect_sel.draw(vis)