Character recognition techniques associate a symbolic identity with the image of character. Character recognition software recognizes text characters in electronic files, usually scanned documents that are saved as images and are not immediately text-searchable. Character recognition software is also known as optical character recognition or OCR, because it uses to optical properties of text to identify the characters and it deals with the recognition of optically processed characters. In this paper, a smart, simple and efficient algorithm is presented, which is mainly designed to identify the region from High Security Number Plates. It is used to identify three regions from vehicle number plates i.e. “Punjab”, “Chandigarh”, and “Haryana”. The proposed algorithm consists of three major parts: Extraction of image, segmentation of characters and the recognition of characters from vehicle number plate. For recognition of characters three features of characters: holes, junction and end points are used. The recognition is based on the number and position of these features.