In traffic management, artificial intelligence is mainly used for moving object detection and recognition. The commonly used application scenarios include the following four parts:
1、 Dynamic illegal evidence collection
Some intelligent transportation systems (ITS) with traffic lenses can use video detection, tracking, recognition and License plate recognition （LPR） to effectively identify illegal behaviors based on image data such as vehicle characteristics and driver’s attitude. Especially for the illegal behaviors that need manual screening, the system can automatically identify the traffic violations in the screen, realize automatic capture and upload, and greatly improve the work efficiency.
2、 Traffic signal control
By using artificial intelligence technology, the traffic flow can be analyzed in real time, the interval between traffic lights can be adjusted, the waiting time of vehicles can be shortened, and the traffic efficiency of urban roads can be improved. Qingdao public security and traffic police department has set up its intelligent transportation system. Through more than 1200 high-definition cameras, the traffic light market of urban main roads, expressways and national and provincial roads has been optimized in real time.
3、 Human vehicle characteristics Association
Human vehicle correlation system, including camera, mobile signal probe and processor, the camera is used to obtain vehicle image information on the highway; mobile signal probe is used to obtain the mobile phone IMEI / IMSI information of passengers on the highway; processor is used to determine whether the vehicle and passengers are in the same position at the same time according to the vehicle driving track and passenger moving track.
4、 Vehicle identification
The vehicle recognition technology based on deep learning extends the feature range from simple license plate or vehicle logo to the whole body. The vehicle’s lights, grilles, windows, etc. are all important features of the vehicle. They can not only identify the brand of the vehicle, but also the sub brand, model, model year and other detailed categories of the vehicle. The retrieval of the designated vehicle in the video image data can also be carried out through local features such as vehicle pictures, annual inspection marks, decorations, etc.
If you need any lens consultant, please contact us