Automatic License Plate Detection of Vehicles with Different Contrast Scenario.

This paper proposes an approach for automatic detection of License Plate if Vehicle in variable contrast scenarios without the use of image thresholding and provides results based on this approach. The results of image thresholding are highly affected by environmental conditions such as sunlight; shade, hue conditions on or around license plate area. Mask images are generated using image morphological operations and are combined together to detect possible license plate candidate areas. These possible candidate detected areas are further probed for presence of license plate characters before identifying it as an actual license plate area. For this probing, contour detection is used, which is later used to categorize clusters of characters inside the detected areas. Once a favorable set of conditions like shape, size, aspect ratio, number of characters are met, the area is confirmed to be a license plate. The proposed algorithm has been tested on images acquired under different test conditions and found to be working successfully.