论文部分内容阅读
在计算机技术和多媒体技术中,扫描仪是图象信息输入的主要输入设备之一,其色彩管理模型是图像在随后复制和传输过程中保持色彩一致性的关键.本文在分析扫描物呈色原理的基础上,提出了一种基于BP神经网络的扫描仪色彩管理模型.首先,该模型对扫描数据进行归一化的预处理,以满足BP网络对数据取值范围的要求;其次,该模型以彩色梯尺区色块代表整个色空间,以减少模型的运算量;再次,模型采用运动量动态调整模型步长,以加快BP网络收敛速度.最后,实验结果表明本文模型提高了色彩转换精度,能够满足扫描仪色彩管理的工程应用需求.
In computer technology and multimedia technology, the scanner is one of the main input devices for image information input, and its color management model is the key to maintain the consistency of color during subsequent copying and transmission.In this paper, , A color management model of scanner based on BP neural network is proposed.Firstly, the model performs normalized preprocessing of the scanned data to meet the requirements of BP network for data range. Secondly, the model In order to reduce the computational complexity of the model, the model uses the amount of motion to dynamically adjust the model step size to speed up the convergence speed of the BP network.Finally, the experimental results show that this model improves the accuracy of color conversion, To meet the color management of the scanner engineering applications.