In this study, the authors propose a real‐time multiple licence plate (LP) detection algorithm for dense traffic conditions which is of vital importance in this modern era due to increased traffic congestion. The chromatic component of YDbDr colour space is proposed to detect the blue regions, whereas a simple yet effective colour detection method is used to identify yellow LP regions. The low‐intensity pixel values are eliminated as a pre‐processing step to enhance the LP regions and Otsu method is used to obtain the binary image. The candidate regions are acquired by using the connected component analysis. The false candidate regions are by large rejected by inspecting the area and aspect ratio of LPs. Additionally, a two‐layered false LP detection approach has been introduced to remove fake LP regions. Experimental results in practical scenarios carried out in various weather conditions show that the proposed method is highly effective in coping with various illumination conditions to accurately detect the multiple vehicle LPs with an accuracy of 93.86%. The average processing time per image is 0.33 s that can achieve real‐time performance.

License plate recognition system (LPR) plays a vital role in intelligent transport systems to build up smart environments. Numerous country specific methods have been proposed successfully for an LPR system, but there is a need to find a generalized solution that is independent of license plate layout. The proposed architecture is comprised of two important LPR stages: (i) License plate character segmentation (LPCS) and (ii) License plate character recognition (LPCR). A foreground polarity detection model is proposed by using a Red-Green-Blue (RGB) channel-based color map in order to segment and recognize the LP characters effectively at both LPCS and LPCR stages respectively. Further, a multi-channel CNN framework with layer aggregation module is proposed to extract deep features, and support vector machine is used to produce target labels. Multi-channel processing with merged features from different-level convolutional layers makes output feature map more expressive. Experimental results show that the proposed method is capable of achieving high recognition rate for multinational vehicles license plates under various illumination conditions.