为了满足实时视频分析中颜色空间转换的效率要求,本研究提出了一种DSP上基于简化移位和查表法(simplified shift and lookup table method,SSLUT)的YCbCr到HSV颜色空间的快速转换算法,并对转换后的图像质量做了综合评价。首先,采用定点移位法把YCbCr转换到RGB,再用查表法将RGB转换到HSV,以实现YCbCr到HSV的快速转换。其次,分别基于DSP和PC两个平台上分析了算法速度。最后,从客观测量和满足人眼视觉特性两个方面,利用峰值信噪比和结构相似性算法综合评价转换后的图像质量。通过算法时间和图像质量评价实验的结果表明,该算法具有良好的实时性。同时,可使图像有较高的保真度。
To meet the efficiency requirement of color space conversion in real-time video systems, a new fast algorithm for the color space YCbCr to HSV conversion was proposed based on the simplified shift and look-up table (SSLUT). The transformed image quality assessment was implemented. The process of color space conversion was divided into two steps. First, the fixed-point shift method was used to convert YCbCr to RGB. Second, the color space RGB to HSV conversion was used for look-up table method. The fast algorithm speed was tested based on DSP and PC Testing Platform. The image quality assessment was achieved using the structural similarity (SSIM) and the peak signal to noise ratio (PSNR). The results showed that the real-time capability was satisfactory and the transformed image had high fidelity.