A Principal Color Clustering Algorithm Based on Audio Feature
Song Na · Journal of Putian University · 2010
The method of extracting principal color is usually split video manually for getting the color test frames,that leads to poor robustness and low automation.By analyzing the feature of sports audio and area color,a principal color clustering algorithm based on audio feature was proposed in this paper.The result of experiment shows that the algorithm don't need artificial pretreatment and have excellent robustness.It improves the extraction accuracy,reduces the time complexity through the audio auxiliary analysis.