LICENSE PLATE DETECTION USING CLUSTER RUN LENGTH SMOOTHING ALGORITHM (CRLSA) Siti Norul Huda Sheikh Abdullah, Marzuki Khalid and Rubiyah Yusof Centre for Artificial Intelligence and Robotics(CAIRO), Faculty of Electrical Engineering, Universiti Teknologi Malaysia, CJalan Semarak,54100 Kuala Lumpur email: mimi@sun1.ftsm.ukm.my,marzuki,rubiyah@utmkl.utm.my Khairuddin Omar Jabatan Sains dan Pengurusan Sistem, Fakulti Teknologi Sains Maklumat, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor. email: ko@ftsm.ukm.my ABSTRACT Vehicle license plate recognition has been intensively stud- ied in many countries. Due to the different types of li- cense plates being used, the requirement of an automatic license plate recognition system is different for each coun- try. In this paper, an automatic license plate recognition system is proposed for Malaysian vehicles with standard license plates based on image processing, clustering, fea- ture extraction and neural networks.The image processing library is developed in-house which referred to as Vision System Development Platform (VSDP). Fixed filter, Mini- mum filter, Median Filter and Homomorphic Filtering are used in image enhancement process. After applying im- age enhancement, the image is segmented using blob anal- ysis, horizontal scan line profiles, clustering and run length smoothing algorithm approach to identify the location of the license plate. Thoroughly each image is transformed into blob objects and its important information such as to- tal of blobs, location, height and width, are being ana- lyzed for the purpose of cluster exercising and choosing the best cluster with winner blobs. Here, new algorithm called Cluster Run Length Smoothing Algorithm (CRLSA) approach was applied to locate the license plate at the right position. CRLSA consisted of two separate new proposed algorithm which applied new edge detector algorithm us- ing 3x3 kernel masks and 128 grayscale offset plus a new way (3D method) to calculate run length smoothing algo- rithm (RLSA), which can improve clustering techniques in segmentation phase. Three separate experiments were performed; Cluster and Threshold value 130 (CT130) and CRLSA with Threshold value 1 (CCT1). From those ex- periments, analysis of error tables based on segmentation errors were constructed. The prototyped system has an ac- curacy more than 96% and suggestions to further improve the system are discussed in this paper pertaining to analysis of the error. KEY WORDS License plate recognition, clustering, run length smoothing algorithm, thresholding. (a) (b) Figure 1. (a)Samples of common Malaysia license plates (b) Samples of special Malaysia license plates. 1 Introduction Automatic license plate recognition system (LPR) is an im- portant area of research due to its many applications. For local authorities license plate recognition is required for the purposes of enforcement, border protection, vehicle thefts, automatic toll collection, and perhaps traffic control. For others, automatic license plate recognition system can be applied to access control in housing areas, automatic park- ing control and marketing tools in large shopping com- plexes, and perhaps for surveillance. Among the commer- cial license plate recognition systems available worldwide are Vehicle License Plate by Oz and Ercal[6] and Cano and Perez-Cortes[6] for real-time environments and Automatic License Plate Recognition using colour by Shi et al.[7] and Kong et al.[3] using texture features and radon transform. In Malaysia, vehicles license plates are in the form of single or double line with normal fonts which com- prise of perhaps 95% of the all the vehicles. Most pictures have been taken in various states in Malaysia like Sabah, Wilayah Perseketuan, Johor, Selangor, Perak, Negeri Sem- bilan, Pahang and Terengganu. There are also special fonts as depicted in Figure 1. LPR normally consists of a camera, illumination, frame grabber, computer, recognition software, hardware (input output adapters) and database as illustrated in Figure 2. LPR employs real time plus artificial intelligence algo- 549-007 323