3.2 Determination of Car License-Plate Location
The windows are opened in the boundaries of candidate regions for the recognition of characters and numbers pattern. To segment the region of each characters , we used horizontal and vertical projection in those boundaries ,for finding charactersinumbers. Then four-layers backpropagation neural networks are performed to classify and confirm the license plate location. The outputs of neural networks are assigned a unique for characterinumbers. Finally , the location and identification of the car license plate is shown on the monitor.