Research And Application Of New Barcode Technology Of Long-distance And Automatic Identification

At present, the use of projector not only shows screen, but also has diverse applications, and shows more charm in the field of human-machine interaction and brings people new experience. Barcode technology is a kind of automatic information collection technology, and it has been widely used in many fields and has played an increasingly important role in the present stages. This paper combines projector and barcode. The bar code technology is applied in projection display and makes users understand the items of information more easily and also make the objects more graphic.What the thesis research on is identify and display the barcode contents in the specific application scenary. The specific application scenary is that the projector at a distance of about 1.2 meters on the desktop projects images and the camera which is at the same level with the projector shoots the screen, and the software scans the projection image to look for barcode and displays contents in the way of projection. Beacause the size of barcode is limited, the image will be reduced, become blur and loss details in the long-distance shot. So serious distortion of common barcodes has been unable to meet the need of the reality.Therefore a new one-dimensional barcode and two-dimensional code based on statistical analysis of global information have been proposed, and they make barcodes more simply and more easily to identify in the captured images. The thesis defines encoding rules of barcodes and reduces the redundant structures, and designs barcodes which are suitable for long distance identification. When decoding new barcodes, the thesis proposes decoding rules which is robust according to encoding rules. In the process of image processing, because the image becomes vague, noised and distorted, the thesis uses Variable threshold processing method based on local image features to select thresholds of binaryzation, and uses adaptive median filter to smooth the image, uses perspective transformation and bilinear interpolation to correct images. When locating barcodes, the thesis uses SIFT to extract feature points and matches images to locate one-dimensional barcode, and the thesis uses position detection pattern to locate two-dimensional code.Traditional recognition systems need to move mobile handheld devices (such as mobile phone) to make barcode into the specified screen in position, and constantly adjust the distance between the hand-held device and the barcode to locate the final realization of optimallocation identification. And barcode recognition system which is designed by this paper, in the process of recognition, don’t need to move the camera or bar code. As long as barcode is in the range of camera captured, using the method based on image processing, the system can recognize barcode automatically. The experiment contrasts and analyzes the results between new barcodes and common barcodes. The result is the new structure of the barcode can be identified accurately in the long distance which becauses its streamlined structure design. So they can meet requirement of the application and they are practical and have popularization value.