A PAPER GAMUT DESCRIPTION MODEL BASED ON BP NEURAL NETWORK
In the paper and printing industry, paper gamut is one of the most important indices of paper color presenting performance. Usually, under some printing conditions, paper gamut polygon is described and formed by six data marks: three of them come from four original solid process ink, such as cyan, magenta and yellow ink, and the other three come from the secondary colors of double print of the three solid processed inks. In spite of the simplicity of the description method in manipulation or test, the polygon formed by the six data marks can’t describe the gamut exactly; because the number of the samples is too small to indicate the feature of the paper gamut perfectly. In the study, based on mapping relation of digital image pixel value and its corresponding printing color, data pairs of digital image pixel values and its corresponding color blocks color chroma values are sampled to establish a printer characterization model based on BP neural network with exactitude to a perfect certainty. Then, more digital image pixel values are sampled at thicker intervals which could be gotten by computer conveniently and promptly, their corresponding printing color chroma values are calculated by the printer characterization model derived above, so the paper gamut is covered by thicker chroma data in the gamut polygon that leads to the best of its ability. Much more polygon vertexes are acquired in far more color hue angel ranges to form the gamut polygon, and the gamut is described much more fully. The experiment shows that the gamut polygon drawn by the printer characterization model describes the paper gamut much more objectively and reflects the paper gamut factually.
paper gamut BP neural network printer characterization model polygon vertices
Zhao Lei Chen Guangxue Tang Baoling
State Key Laboratory of Pulp and Paper Engineering,South China University of Technology, Guangzhou,5 State Key Laboratory of Pulp and Paper Engineering,South China University of Technology, Guangzhou,5
国际会议
广州
英文
1771-1775
2010-11-08（万方平台首次上网日期，不代表论文的发表时间）