1. Background
Illegal parking of vehicles at roadsides for loading and unloading activities is often one of the main causes of traffic congestion.
To maintain smooth flow of traffic, a series of traffic management measures have been implemented along the section of Queen's Road Central between Ice House Street and Pedder Street. These include no-stopping restriction (NSR) at any time and NSR by time period or vehicle types, box junctions and yellow striped light crossing. However, illegal parking of vehicles at the above road section has created traffic bottle necks and triggered traffic congestions. Similarly, vehicles illegally parked along the NSR at Hoi Yuen Road near the roundabout at Kwun Tong Road has created congestions at Hoi Yuen Road, and even affecting the traffic at the roundabout, Kwun Tong Road and Hip Wo Street.
 |
 |
| Queen's Road Central |
Hoi Yuen Road |
Albeit police officers' continued patrolling and enforcement efforts at the scenes to clear the traffic, the above traffic offences and congestions will recur soon after the police officers leave the scenes. Making use of the development in artificial intelligence and sensing technologies, the government intends to make use of these technologies to carry out automatic traffic enforcement, strengthen the deterrent effect to drivers for the above illegal acts, and reduce road congestions.
Therefore, the Transport Department (TD) in collaboration with the Police is conducting testing of automatic traffic enforcement system (the system) at the above two road sections. The system aims at identifying and recording common roadside traffic offences such as illegal stopping at road sections with NSR, which have given rise to traffic congestions.
2. Components of the System
The system comprises CCTV cameras and associated equipment (including computer, battery and conduits) installed at suitable lamp posts at the road sections concerned for traffic surveillance. The system is energized by the street lighting system so that it can operate around the clock.
3. Operating Principle of the System
The system utilises video analytics, artificial intelligence and vehicle licence plate identification technologies for real-time analysis of vehicle movements and recording of roadside traffic offences and the licence numbers of offending vehicles. Through deep learning of the system under varying environment (e.g. daytime and night time light conditions, weather conditions), traffic signals and road conditions (including pedestrian crossing), the system will be able to accurately analyse the vehicle movements at the road sections in real time and to identify traffic offences of the vehicles. The system will output relevant traffic offence data and evidence to facilitate future enforcement.
During the testing period, the videos and data collected will be used for training and testing of the system, and for optimising the system's settings and evaluation of its effectiveness.
4. Privacy Protection
The system will strictly comply with the requirements of the Personal Data (Privacy) Ordinance. The CCTV cameras will be adjusted to cover the road sections concerned. The system has no facial recognition function and will only identify licence plate numbers. If no offences are identified, the video records concerned will be immediately deleted from the onsite equipment. Only video records capturing offences will be output. All data collected will be encrypted.
5. Testing Locations
Queen's Road Central near Pedder Street and Ice House Street:
Hoi Yuen Road near its Roundabout to Kwun Tong Road:
Banners and notices were put up at appropriate locations of the above road sections during the testing to remind drivers that the automatic traffic enforcement system is under testing. The TD is reviewing the effectiveness of the system and is discussing with the Police on the way forward for automatic traffic enforcement.
Testing on Automatic Traffic Enforcement System
