Reasearch On Automatic Acquisition And Recognition System Of Train Carriage License Image

Train license management plays a vital role in train management system. To improve the manual management of train license, this thesis proposes the automatic management methods based on the image processing technology. It is mainly about the acquisition, match and mosaic, selection of the meaning image and train license positioning. It is significantly for train amending and dispatching management.For image collection, we use motion detection algorithm to detect train movement so as to fulfill the automatic acquisition of motion images. This paper proposes a mixed algorithm that combining temporal difference and background illumination adjustment. We detect the moving object (train) through its brightness differences value. The experimental results show that this algorithm can satisfy the requirement of real-time detection of train license.For the image matching and image mosaic, the main purpose is to stitch the images of incomplete train license and fulfill the task of train license plate location .This paper deeply studies the image matching algorithm based on feature extraction of SIFT and the matching algorithm based on edge extraction, and verifies the image Mosaic algorithm. Finally, stitch the incomplete images.For binary image, we propose a binary image threshold algorithm based on the traditional Otsu algorithm and the adaptive background. Firstly, the Otsu threshold value is used to segment background and object in image; secondly, changing the threshold according to the brightness of the image. Experiments show that the algorithm is simple in design and well meets the requirements of real-time system.According to the study in the location of car plate license ,we propose a novel method to choose the interested region based on contour features. Classify the plate and determine whether exists in the given image. Experiments show that the algorithm meets the system requirements effectively.