Obstacle Detection Using Combination Between Gradient Analysis and Threshold Process

Wheelchair is very important for disability person. Currently a smart wheelchair has been developed that can move autonomously. Generally vision systems are used to control Smart Wheelchair autonomously. Detecting obstacles is one of the important parts of autonomous Smart Wheelchair. Obstacle detection can give an information for Smart Wheelchair to take a decision. The problem of obstacles detection using camera is a shadow. We uses a combination between Prewitt and threshold to detect an obstacle. The Prewitt operation was used for detecting edge of object in image. However, sometimes an object in the image contains shadow. Furthermore, the elimination of shadow object in an image using threshold process. Result of accuracy on in this research is 80.36% from the 8 data testing. Later, the smart wheelchair can decide the right obstacle and the left obstacle, if the obstacle in the right of wheelchair, wheelchair should be turn left.