STEREO IMAGE PROCESSING FOR DISTANCE AND SIZE MEASUREMENT

The stereo vision consist of two cameras that provide two image from different angle and coordinate. The stereo camera will represent as human eyes which have capability on estimation the distance and the size of the object. Most of current stereo image processing will require the same stereo camera for calibration and image processing which limit the usage and performance for stereo image processing in various field. Cross stereo camera platform will be another option for stereo image processing as it will able to process any source of stereo image. The understanding the capability, setup and configuration is basic or fundamental for stereo image processing also it will affect the end result. Develop a suitable algorithm for analysing stereo image processing for distance and size measurement is important also the limitation of the stereo image processing. The result from this project using different source of images for calibration and image processing provide a 65% of distance and size measurement accuracy which is not good enough since the result is not precise due to different camera parameter produce by the calibration stage such as focal length. To improve this limitation and accuracy, a new calibration and detection algorithm that can process cross cameras images