Fuzzy based hand-shake compensation for image stabilization

The paper proposed a way of dealing with optical image stabilization in solving blurring images caused by hand shake. The idea is to determine the hand-shake situation and then to correct blurring image through position compensation. This method directly detects motion signals to distinguish the hand shake situations from normal camera movement. In the process, fuzzy rule mechanism is employed to have more accurate decision. If a hand-shake situation is determined, the corresponding correction signal is generated to correct the image in a real-time fashion. In order to demonstrate the effectiveness of the proposed approach, in our implementation, the system directly moves the camera mounted on an X-Y platform to compensate the hand shake effect. From our experiments, it is clearly evident that this method indeed can have better image quality. The average shake without using this method is 73.8 pixels, the average shake with this method is 33.4667 pixels, the average shake reduction is 40.333 pixels and the average percentage of shake reduction is 54.64%.