Vehicle License Plate Recognition (VLPR)

In general, human can easily read the character but the computer cannot read the character easily unless the computer is pre-trained to do so. VLPR deals with License Plate localization and vehicle number recognition. VLPR has three modules namely License Plate Extraction, Segmentation and Recognition. Extraction of plate is mostly carried out using edge detection. In Segmentation various tasks such as Filtering, Thinning, Vertical and Horizontal projections are performed. Recognition is the final stage to recognize character and numbers. The fulfillment of the Automated License Plate recognition system depends on quality of the input image. The performance of VLPR is degraded due to noise such as shadow, occlusion, rain streaks and blur due to vehicle speed. This paper analyzes various VLPR methods, classifies the VLPR literature based on feature extraction and classification schemes, presents the survey of approaches for detection, localization and recognition of plate character & numbers and addresses the challenges in performance improvement. Based on the challenges, an efficient scheme is suggested for VLPR which can handle dynamic weather condition.