ROBUST METHOD OF SEGMENTING CHARACTERS OF LICENSE PLATE ON IRREGULAR ILLUMINATION CONDITION

摘要

PURPOSE: A key pad character segmentation method in irregular lighting environment is provided to obtain weighted value by Chi square probability density function and increase the brightness and contrast of an image and obtain high quality of binarized image, thereby accurately extracting a character of key pad. CONSTITUTION: A number plate candidate area is extracted from a vehicle image(S2). A weighted value is obtained by Chi square probability density function. A new pixel value is obtained by the weighted value(S3). The rotation of the number plate image is determined(S4). The number plate image is corrected by rotation converting(S5). The phase relation and location between characters in number plate area and the size of the number plate are investigated in advance by shape information of the number plate. A character is extracted from the number plate image by vertical/horizontal profiling(S6~S10).