Abstract: An efficient reintegration of the disabled people in the
family and society should be fulfilled; hence it is strongly needful to assist their diminished functions or to replace the totally lost
functions. Assistive technology helps in neutralizing the impairment.
Recent advancements in embedded systems have opened up a vast
area of research and development for affordable and portable assistive devices for the visually impaired. Granted there are many assistive devices on the market that are able to detect obstacles, and numerous research and development currently in process to
alleviate the cause, unfortunately the cost of devices, size of
devices, intrusiveness and higher learning curve prevents the visually impaired from taking advantage of available devices. This
project aims at the design and implementation of a detachable unit
which is robust, low cost and user friendly, thus, trying to
aggrandize the functionality of the existing white cane, to concede above-knee obstacle detection. The designed obstruction detector
uses ultrasound sensors for detecting the obstructions before direct contact. It bestows haptic feedback to the user in accordance with the position of the obstacle.
Abstract: This paper presents a new classification algorithm using colour and texture for obstacle detection. Colour information is computationally cheap to learn and process. However in many cases, colour alone does not provide enough information for classification. Texture information can improve classification performance but usually comes at an expensive cost. Our algorithm uses both colour and texture features but texture is only needed when colour is unreliable. During the training stage, texture features are learned specifically to improve the performance of a colour classifier. The algorithm learns a set of simple texture features and only the most effective features are used in the classification stage. Therefore our algorithm has a very good classification rate while is still fast enough to run on a limited computer platform. The proposed algorithm was tested with a challenging outdoor image set. Test result shows the algorithm achieves a much better trade-off between classification performance and efficiency than a typical colour classifier.