The article presents the results of experiments related to the use of super-resolution technology to improve the image quality used for recognizing license plate numbers of moving vehicles. An acquisition system based on two cameras rotated relative to each other was proposed, which resulted in improved readability of license plate numbers after applying the super-resolution algorithm. The obtained results were compared with the effects of using the same total number of frames recorded by a single camera or two cameras without rotation around the optical axis. The proposed method can be used in Intelligent Transportation Systems to improve the reliability of vehicle identification at distances from the camera using image analysis techniques.APPLICATION OF THE TWO-CAMERAS VIDEO SEQUENCE ACQUISITION FOR THE VEHICLES’ REGISTER PLATES’ NUMBER RECOGNITION WITH THE USE OF THE SUPERRESOLUTION TECHNOLOGY
In the paper some experimental results are presented, which are related to the use of the superresolution technology for the quality improvement of the image used for further register plates’ numbers recognition of the moving vehicles. Proposed video acquisition system is based on two-cameras rotated relatively to each other, which resulted in the improvement of the improving the readability of plates’ numbers after the application of the superresolution algorithm. Obtained results have been compared with the effects of applying the same total number of frames acquired by a single camera or two cameras with no rotation around the optical axis. The method proposed in the paper can be used in
Intelligent Transportation Systems to improve the reliability of identification of vehicles in long distances from the camera performed using image analysis techniques. (…)
The article contains 16,009 characters.
Source: Logistyka Magazine