Robust Chinese license plate recognition method in non-controllable environment

• Main Authors: XU YONGCUN, SUN HAO, AN XIN, LI HONGFANG, SHAN ZHENG, FAN DONGNAN, BAI XUEFENG, ZHUO LI, AN ZEPING, LIANG ZHAOWEI
• Format: Patent
• Language: Chinese; English
• Published: 17.06.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

The invention discloses a robust Chinese license plate recognition method in a non-controllable environment, belongs to the field of image processing, and aims to solve the problems of poor robustness, low performance and the like in license plate recognition in the non-controllable environment. The method is realized based on a CRNN framework, and comprises three steps of establishing a data set, preprocessing an input image and designing a network structure. As the lengths of the Chinese license plate characters are different, the CTC algorithm is adopted to solve the alignment problem of the variable-length license plate characters. The CTC can convert a series of character sequences obtained from the sequence recognition layer into a final character sequence. The method has the advantages that the recognition speed is high, and the accuracy is high; 2, recognition of any fixed-length license plate character sequence can be realized, and recognition requirements of various license plates with different len