| Along with the development of the information technology, the barcode technology is getting more and more important in the informationization industry, and the application is getting more and more widespread as well. In China, in the traceable system of the animal epidemic prevention, the animal earmark 2D barcode as an information storage marking has recorded the documents an animal from the birth to the death, including the birth, the immunization injection, the quarantine, the transportation, slaughtering, etc. Therefore, the research of recognition the animal earmark barcode plays a very important role in the animal epidemic prevention and the animal food source security.The article through the analysis on the ear mark image of the pig from the camera, based on the analytic thoughts of the energy histogram that has completed the designing and realizing of the barcode coding-area gaining algorithm, and carrying on the image processing to the coding-area image, to achieve the recognition of the 2D barcode.The article first through the analysis on the earmark image characteristics which gathered under three different section by the system, choosing the difference value on the image line and row direction separately as the characteristic factor, combing respective characteristics of three sections of images, to achieve the algorithm of obtaining barcode coding-area under complicated background, which based on image energy histogram by the system. Tests show that the method can obtain accurate image of the code.Secondly, according to the comparison and the analysis to the common image pretreatment algorithm, the particular 2D barcode image processing algorithm has designed to the system. Aiming at the illumination uneven images, the article has used the auto-adapted binaryzation method, which has solved the problem of the terrible binaryzation effect due to the illumination reason. The algorithm use the weighted average on the picture element surrounding the target point as a threshold value to carry on point by point the binaryzation to the image, it shows simple computation, fast division speed and strong compatibility.Finally, in order to locate the barcode boundary simultaneously to guarantee high efficiency of the algorithm, the system uses the Hough algorithm to liner fit the coding-area boundary, the union mathematics characteristic, has determined the coding-area vertex coordinate has determined by gathering the characteristic of the intersection straight line. The geometry correction has performed through the vertex coordinate on the barcode image, and then the image would be sampling adjusted to complete to the barcode image recognition, and 2D barcode is conversed into a matrix composed with"0"and"1".By adopting modular programming design, the ear mark image recognition software of the pig is realized on PC. |
PDF Full Text Request