Character Translation on Plate Recognition with Intelligence Approaches

In recent years, the number of automobiles in Indonesia has expanded. This rise has a knock-on impact on street crime. On this problem based, a preventative road safety prevention system is required. This research contribution is to develop an efficient algorithm for detecting vehicle license plates. This study's technique incorporates artificial intelligence technology with character translation. Yolov3 and Yolov4 are the artificial intelligence systems employed in this study. The detection of objects in the form of license plates is the result of this approach. In artificial intelligence, object detection results are utilized as input for image processing. The image processing method is used to translate characters. Optical Character Recognition (OCR) is used to decode the characters in the image precisely. The artificial intelligence data training resulted in a 76.53% and 89.55% mean average precision (mAP) level. Using OCR, the system is capable of character translation. These results give an opportunity to develop more complex image-processing applications.