Brazilian vehicle identification using a new embedded plate recognition system

• Novel embedded plate recognition system.
• A robust and powerful approach to brazilian vehicle recognition and classification.
• Analysis of the impact of the image preprocessing methods.
• Proposed method has very good results in terms of efficiency and effectiveness.

Expert systems for parking lot access control are developed in vehicle management through tracking and number recognition. These systems commonly use cameras to identify a vehicle through its license plates based on intelligent and optical character recognition techniques. This paper presents a new system to detect and recognize Brazilian vehicle license plates, in which the registered users have permission to enter the location. For this, techniques of Digital Image Processing were used, such as Hough Transform, Morphology, Threshold and Canny Edge Detector to extract characters, as well as Least Squares, Least Mean Squares, Extreme Learning Machine, and Neural Network Multilayer Perceptron to identify the numbers and letters. The system was tested with 700 videos with a resolution of 640 × 480 pixels and AVI format, granting access only when the plate was registered, getting a 98.5% success rate on the tested cases. The movement detection step is linked to the system, becoming faster and more accurate in real time. Thus, can be concluded that the proposed system is a promising tool with high potential which can be applied commercially.