Intrusion detection based on computer vision technology realizes real-time observation of a dynamic scene by designing an image processing method, and sends the intrusion detection results to the upper management system when there is an external intrusion in the scene; it is required to independently write the program source code with the following functional modules:
1. 2. Continuously observe a dynamic scene through the built-in camera of the mobile phone/personal notebook
; 2. Acquire the real-time video stream data of the camera, and save the picture in the form of jpg/bmp/png at a certain time interval;
3. 3. At least one kind of preprocessing is performed on the image, such as Gaussian smoothing, histogram equalization, etc
.; 4. When there is intrusion in the scene, detect the intrusion object and display the image of the intrusion object;
5. 5. Display intrusion objects in at least three different characteristic forms, such as Voronoi diagrams, edges, corners, etc
.; Display the moving track of the object on the image;
6. 7. Predict the moving speed of objects with an accuracy of not less than 60%
; The attribute information of the object, such as shape, size, moving speed, etc., is simulated to the upper management system.
Effect picture:
Effect video:
python opencv vehicle speed measurement video car speed detection intrusion detection speed measurement
Download link: