Mean Shift-Based Target Tracking for Moving Spherical Object in Video
Autor: | Bo Zhao Li |
---|---|
Rok vydání: | 2014 |
Předmět: |
Color histogram
Color normalization business.industry ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Optical flow Pattern recognition General Medicine Feature (computer vision) Histogram RGB color model Computer vision Mean-shift Artificial intelligence business Histogram equalization Mathematics |
Zdroj: | Applied Mechanics and Materials. 623:156-160 |
ISSN: | 1662-7482 |
Popis: | When researching on target tracking in the on or off line video, using a variety of methods such as MeanShift, Camshaft’s, feature points and optical flow algorithm. MeanShift target tracking algorithm is introduced in this paper. Firstly, tracking object is selected by human-computer interaction. Then color feature histogram is obtained using RGB color information, and color distribution probability image is got by converting color feature histogram. Finally, by comparing the probability difference of color distribution of the adjacent frames, motion directions of the object’s center are obtained, which object can be effectively tracked. |
Databáze: | OpenAIRE |
Externí odkaz: |