Controlling Mobile Robot Navigation Equipped with Computer Vision Perception for Text using OCR Technique

Recently, computer vision technology has been widely used to understand the visual world by extracting information depending on the digital images, which is done by using a conventional camera with the designed system. This work proposed a system that will employ computer vision for text detection by using the Optical Character Recognition (OCR) technique, where the camera is supported on the four-wheel derived mobile robot. The proposed system integrates the mobile robot navigation and the functionality of OCR to detect the signs text by scanning the street signs, the function of perception text system consists of two steps first one building four-wheel mobile robot using Arduino UNO microcontroller which represent an opensource software and hardware Italian company, UNO meaning one in Italian language, then navigate this mobile robot by using HG-05 Bluetooth wireless communication module, the second step represents the process of detecting text inside images captured by web camera install on the mobile robot and converting the text into an electronic form, the two steps realized by OpenCV(Open Source Computer Vision Library) python by importing cv2 method, using tesseract library, and ImageGrab module in python to detect and recognition text, the proposed system can be employed the digitization to reduce the effort of humans in battle zoon and storing these information in dedicated documents, the proposed system can be employed for military applications, detect path in battle zoon by detection and recognition the text plate in that zoon. The perception text method is presented by using the OCR technique and an android phone with a web camera, where all this equipment is connected to the mobile robot, the result will demonstrate the text extracts match 100% with the snapped image from real-time video, by transforming each road text plate to save as a text file containing a stream of information represent the corresponding character from the real-time video to the road sign plate, and this text file can be saved as a PDF.