Skip to content

Add object detection, tracking, mobile notifications, and search to any security camera.

clean

roryclearcc6bedf27114
diff --git a/test/test_mot.py b/test/test_mot.py
index 434343d..fa3fba3 100644
--- a/test/test_mot.py
+++ b/test/test_mot.py
@@ -9,7 +9,7 @@ from pathlib import Path
@TinyJit
def do_inf(im, model): return model(im)
-def letterbox(im, new_shape=(1280, 1280), color=(114, 114, 114), auto=True, scaleFill=False, scaleup=True, stride=32):
+def letterbox(im, new_shape=(960, 960), color=(114, 114, 114), auto=True, scaleFill=False, scaleup=True, stride=32):
shape = im.shape[:2]
if isinstance(new_shape, int): new_shape = (new_shape, new_shape)
r = min(new_shape[0] / shape[0], new_shape[1] / shape[1])
@@ -34,6 +34,12 @@ def letterbox(im, new_shape=(1280, 1280), color=(114, 114, 114), auto=True, scal
im = cv2.copyMakeBorder(im, top, bottom, left, right, cv2.BORDER_CONSTANT, value=color)
return im, ratio, (dw, dh)
+def scale_coords(boxes, ratio, dwdh):
+ boxes[:, [0, 2]] -= dwdh[0]
+ boxes[:, [1, 3]] -= dwdh[1]
+ boxes[:, :4] /= ratio[0]
+ return boxes
+
if __name__ == "__main__":
from ocsort_tracker import ocsort
ocs_tracker = ocsort.OCSort(max_age=60)
@@ -59,7 +65,7 @@ if __name__ == "__main__":
while True:
ret, im0 = cap.read()
if not ret: break
- im = letterbox(im0, new_shape=(960, 960), stride=32, auto=True)[0]
+ im, ratio, (dw, dh) = letterbox(im0, new_shape=(960, 960), stride=32, auto=True)
im = im.transpose((2, 0, 1))[::-1]
im = np.ascontiguousarray(im)
im = Tensor(im).cast(dtypes.float32)
@@ -67,19 +73,20 @@ if __name__ == "__main__":
if len(im.shape) == 3:
im = im[None]
pred = do_inf(im, model).numpy()[0]
-
+ pred[:, :4] = scale_coords(pred[:, :4], ratio, (dw, dh))
# no tracker, todo clean
#pred = pred[pred[:, 4] >= 0.25]
#pred = rescale_bounding_boxes(pred, from_size=(im.shape[2:][::-1]), to_size=im0.shape[:2][::-1])
#_, buffer = cv2.imencode(".jpg", im0)
#out.write(draw_bounding_boxes(buffer, pred, class_labels))
- online_targets = t.update(pred, [960,960], [960,960], 0.25)
+ h0, w0 = im0.shape[:2]
+ online_targets = t.update(pred, [w0, h0], [w0, h0], 0.25)
preds = []
for x in online_targets:
if x.tracklet_len < 1 or x.speed < 2.5: continue
if x.class_id == 0 and x.track_id not in ppl: ppl.add(x.track_id)
- preds.append(np.array([x.tlwh[0],x.tlwh[1],(x.tlwh[0]+x.tlwh[2]),(x.tlwh[1]+x.tlwh[3]),x.track_id,x.class_id,x.score]))
+ preds.append(np.array([x.tlwh[0], x.tlwh[1], x.tlwh[0] + x.tlwh[2], x.tlwh[1] + x.tlwh[3], x.score, x.class_id]))
#tlx tly w h, track_id, age, class_id, score
print("ppl =",len(ppl))
_, buffer = cv2.imencode(".jpg", im0)