The proposed system is a camera-based assistive text reading framework to help blind persons detect currency and identify the obstacle in front in addition to read text labels and product packaging from hand-held objects . To identify text from cluttered backdrops, the image is first converted to gray, then to binary form. Text is localized using text localization algorithm and haar cascade is employed to identify the text and then e-speak engine converts the text into voice output. The above system is enhanced to recognize any Obstacle in front and produce the voice output through Ear phone to blind users. Adaboost model is employed in the obstacle detection process wherein it identifies human and vehicles apart and is given as voice output. To develop a novel camera-based computer vision technology to automatically recognize banknotes to assist visually impaired people. A novel component-based banknote recognition system by using SURF (Speeded Up Robust Features) to achieve high recognition accuracy and to handle various challenging conditions in real world environments is proposed. The input image is converted to gray, from which the descriptors and key points of note are extracted and compared with template and the voice output is produced through Earphone to blind users.