earticle

영어

Tracking objects in real time and exact tracking have long been challenging area in the field of computer vision. This paper describes a modified Continuously Adaptive Mean Shift (CAMshift) algorithm based on the Hue Saturation Value (HSV) color model for tracking an object in real time. The existing CAMshift can detect precisely when an object has a simple color. However, it has some disadvantages. When a CAMshift algorithm tracks an object, it does not consider direction and velocity, and it is difficult to track an object that has various colors and when the background contains color(s) similar to those of an object. In such case, CAMshift is worse than Meanshift. To overcome these disadvantages, a Kalman filter is used with CAMshift, as a Kalman filter can obtain direction vectors and it can make CAMshift exact, even under the aforementioned conditions.