Distance alert system using Stereo vision and feature extraction

Abstract

In the recent years', image processing is becoming an important field in computer science. Also, interaction between human and computer is in increasing rate. Nowadays image processing used in many application for example application like measuring the distance between systems that consist of a camera and objects around the system using Stereo vision. Some previous studies use pixel by pixel search for object detection this sometimes leads to failure in detecting object, also some sensors like radar used for measuring distance get all distance for multiple object but it presents them in 2 Dimension (2 D) grid. Above of that sensors measuring system are expensive and power consuming. The objectives of this paper are developed stereo vision system to measure the distance of detected object without the need for pixel by pixel search methodology. The methodology is, the system implement equation for finding the real distance of detected objects, this detection done by feature extraction method, then the system fire alarm if the detected object being in critical distance from the system. In conclusion, the system can be applicable in a wide real life applications because of its high accuracy, low power consumes, cost effective and easy to be embedded with other systems such as navigation systems in robotic system.