Once we have detected the License Plate we have to crop it out and save it as a new image. This gets trickier if the image does not even have a car, in this case we will an additional step to detect the car and then the license plate. Normally the detection algorithm is trained based on the position of camera and type of number plate used in that particular country. The accuracy can be improved if we know the exact size, color and approximate location of the number plate. We will use the contour option in OpenCV to detect for rectangular objects to find the number plate. The first step is to detect the License plate from the car.
model.predict() Now we are detecting number plate. A test was carried out on LP images with varied angles of orientation and the accuracy rate of line and character recognition were 96.94% and 91.76% respectively.And also we predicted the model of that image using. The models from these trainings were deployed with the whole license plate testing.
Differently from a one-line type license plate, the research proposed an approach of two-step training and recognizing by firstly segmenting and training the 3 lines of each license plate by using deep learning (MobileNets and Inception-v3) in the second step, Thai letters and numbers on the plate were segmented and trained. This type of plate is a triple-line type containing 3 lines of character sets, which represent category, registered place, and registered numbers respectively from top to bottom. This research focuses on the latter part by researching into Thai motorcycle license plate. The vehicle license recognition system commonly combines 2 sub-systems: license plate detection, which aims to locate the vehicle and its license plate and license plate recognition, which aims to recognize the characters on the plate. Recognizing vehicle’s license plate is necessary because the number of vehicles is increasing and it goes beyond human’s ability to complete this task.
0 kommentar(er)
