Trial license plate image processing all 10 shots were found to try first is to separate the letters and numbers from the default license plate by plate, a two-level (Binary Image) and snip out a specific portion images cut into groups, each group of white pixels, which is pixel event group for each group of letters and numbers on the plate, plate. The result can be split out every letter a. The next step is to divide the categories group letters and numbers to letters, the province using the height of the letters and numbers in pixels. Then each went to THA na, finding the edge of an image using the method of the solution Bell when na result from programming and use of the program came to Visual comparison. Results from the edge programming thickness over the use of the program. In the next step is to find out the core structure of the image (the Skeleton) resulting when Panama compared to the use of the program. I found that there is no difference in the image-edge exactly and stored the coordinates point to na x axis and y axis.
การแปล กรุณารอสักครู่..