摘要
相机的成像原理决定了它不像人眼一样有很强的适应性,对于一些亮暗分明的场景,只能取亮或暗的部分而无法兼顾。高动态范围图像将一组亮暗程度不同的图像,取出它们的优质区域,然后合成到一起,从而兼顾了场景的亮暗部分。合成HDR图像首先需分别标定R、G、B通道相机的响应曲线,然后通过该曲线可得到图像对应通道每个像素点的照度值,再使用色调映射算法就可将像素点照度值转换成灰度值显示在屏幕上。然而在RGB空间下传统的色调映射算法会压缩R、G、B通道的颜色差异,导致图像颜色失真,由此提出了一种色彩调节的算法,该算法首先提取原始序列图的颜色控制系数,然后用该系数调整色调映射算法产生的颜色差异,从而使合成图的颜色与序列图保持一致。
The principle of Camera imaging makes it can't have the strong adaptability as human eyes, for some scenes which have a distinct set of tonal steps from light to shadow,normally take either bright or dark parts. In dealing with a set of images of different brightness degree, high dynamic range images take out their high quality region and combine them together, so as to take into account the light and dark of scenes. Synthetic HDR images first need calibration channels R,G and B separately of camera response curve,and through this curve we can get their luminance values for each pixel of the image for corresponding channel, and then pixel luminance value can be converted to gray scale values and displayed on the screen by using the tone mapping algorithm. However,in the RGB colorspace,the traditional tone mapping algorithm will compresses the color difference of R,G,B channels,result in color distortion, thus we proposed a color adjustment algorithm which extract color control coefficients of the original sequence firstly, then adjusting color differences of tone mapping algorithms through the coefficients, so that the color of the composite image could keep pace with the sequence diagram.
出处
《长春理工大学学报(自然科学版)》
2015年第5期145-149,共5页
Journal of Changchun University of Science and Technology(Natural Science Edition)
关键词
动态范围
响应曲线
照度值
色调映射
色彩调节
dynamic range
response curve
luminance values
tone mapping
color adjustment