1、1.video jitter problem description2. Video jitter detection principle3. Video jitter detection implement4. Video jitter detection result5. ConclusionVideo jitter detection 1. happen in the traffic surveillance video 2. Cameras exposure to environmental disturbances 3. Cameras are being moved by some

2、body 4. Reduces the effectiveness of surveillance videovideo jitter problem description 1.In the normal video, the points in the current video frame will not move significantly in the next frame. 2.In the jitter video, the points in the current video frame will move significantly in the next frame.V

3、ideo jitter detection principle1. Selecting feature points2. Feature points matching3. Calculating proportion4. Current video frame jitter judging5. Video jitter judgingVideo jitter detection implementalgorithm process framework1.normal video detection result 2.jitter video detection resultVideo jit

4、ter detection resultthe number of feature point: 299the radius of the searching circle: 2.5the number of points in the match: 250the match proportion: 250/299=0.84 the threshold value of match proportion: 0.70current video frame jitter judging: no jitterthe number of detect frame: 100the number of j

5、itter frame: 5the threshold value of jitter frame: 60Video jitter judging: no jitternormal video detection result the number of feature point: 299the radius of the searching circle: 2.5the number of points in the match: 0the match proportion: 0/299=0 the threshold value of match proportion: 0.70current video frame jitter judging: jitterthe number of detect frame: 100the number of jitter frame: 83the threshold value of jitter frame: 60Video jitter judging: jitterjitter video detection result1.the feature point matching2.high detection accuracy3.high detect