Car plate recognition based on UMACE filter

摘要

In this work, the Automatic Parking and Retrieval System (APRS) is proposed to overcome the parking space problem. This system will save space, time, environment and security. The main part that involves in this system is car plate identification and recognition. The plate recognition is used as an identity number to park the car into the available parking space or slot. Therefore in this paper, car plate recognition based on Unconstrained Minimum Average Correlation Energy (UMACE) filter is discussed. Peak to Side lobe Ratio (PSR) is used as a performance measure via the sharpness of the correlation peak. The implementation involved only two stages, as compared to conventional method that consisted of segmentation stage of character. Over hundred images are used as database to evaluate the method and results showed that the proposed method is able to classify car plate with good accuracy.