This paper proposed a fasfalgorithm for YCbCr space to HSV space conversion based on the analysis of the algorithms between YCbCr to RGB and RGB to HSV conversions. Rather than using the floating point multiplication, it used the shift operator and look-up table method respectively in this algorithm, so that improved the efficiency on fixed-point DSPs significantly. In addition, Y component was not involved in the calculation during the conversion of the algorithm, thereby further reduced the computational complexity . Finally,according to the experiments, it obtained the consequence that this algorithm could save 80% of the computing time than the traditional algorithm on fixed-point DSP platform and 46% of the computing time on PC platform. Therefore, the proposed algorithm can be used in high real-time video analysis systems, such as license plate recognition and flame detection.%通过分析YCbCr到RGB以及RGB到HSV之间转换的算法,提出一种YCbCr空间转换到HSV空间的快速算法.在该算法中分别使用了移位运算和查表法代替了浮点乘法运算,从而显著提高了算法在DSP上的运行速度.另外,转换时Y分量不再参与计算,从而进一步降低了运算复杂度.最后实验证明,在DSP平台上,该算法比传统算法能节省80％的计算时间,在PC平台上能节省46％的计算时间.因此,提出的算法在车牌识别、火焰检测等实时视频分析应用中有广泛的应用性.
展开▼
