Implementation Of Object Detection and Identification For Visually Impaired

The issue of visual debilitation or visual deficiency is confronted around the world. As per insights of the World Health Organization (WHO), worldwide, somewhere around 2.2 billion individuals have a dream weakness or visual deficiency, of whom no less than 1 billion are visually impaired. As far as territorial contrasts, the pervasiveness of vision disability in low-and center pay districts is multiple times higher than in big league salary locales. Blind individuals by and large need to depend on white sticks, guide canines, screen-understanding programming, magnifiers, and glasses to help them for versatility, be that as it may, to help the visually impaired individuals the visual world must be changed into the sound world with the possibility to illuminate them about objects as well as their spatial areas. In this manner, we propose to help the outwardly debilitated by presenting a framework that is generally possible, smaller, and financially savvy. The goal of the project is to develop a model an object recognition model to detect and recognize the object for visually impaired people and it can also be used in other markets. Existing techniques for object detection requires a large training data it takes more time and the process is complicated. In this project we propose computer vision concept to recognize object and convert it into text by using the pre trained dataset and the texts are converted further to speech. The system will also support multiple object detection and building a one line description of the object which can be either describe the position of the object or the color. The proposed system can further be implemented in any portable devices for the visually impaired people to recognize objects.