论文部分内容阅读
提出了一种新的阴极射线管特性化的方法。该方法的特点是采用“视觉匹配”方法 ,在反射体表面色和自发光体之间映射一些色貌因素 ,但没有使用任何复杂的色貌模型。是一种考虑了一些色貌因素的阴极射线管特性化方法。由于该问题的个性因素较多 ,采用BP神经网络实现色空间的非线性映射。实验结果表明 ,只要阴极射线管被标定 ,在办公室环境下 ,该方法可以改进在不同的阴极射线管上重现的颜色。采用 3 7 7 3简单的网络结构 ;分色相样本训练。训练样本平均色差可以达到 3.0 7L u v 色差单位 ,测试样本平均色差可以达到 4 .5 5L u v 色差单位 ,小于阴极射线管的最大可接受色差 ,结果是令人满意的。这在电子商务和办公自动化方面有广泛的用途。
A new method of characterization of cathode ray tube is proposed. The method is characterized by the use of a “visual match” method that maps some color appearance factors between the surface color of the reflector and the self-luminous body but does not use any complicated color appearance model. Is a cathode ray tube characterization method that takes into account a number of color appearance factors. Due to the personality factors of this problem, BP neural network is used to realize the nonlinear mapping of color space. The experimental results show that this method can improve the color reproduction on different cathode ray tubes as long as the cathode ray tube is calibrated in the office environment. Adopt a 3 7 7 3 simple network structure; dichroic sample training. The average color difference of training samples can reach 3.0 7L u v color difference unit, the average color difference of test sample can reach 4.55 uv color difference unit, less than the maximum acceptable color difference of cathode ray tube, the result is satisfactory. This has a wide range of uses in e-commerce and office automation.