基于BP算法的纹理图象合成研究

针对目前纹理图象合成技术存在的弱点，应用人工神经网络BP算法实现了纹理图象的合成。通过对纹理图象的分析和特征提取来获取图象的各种参数信息，然后采用人工神经网络BP算法，建立BP网络模型，将原始图象的特征函数作为BP网络的输入，通过学习-训练，确定该图象的人工神经网络结构，并最终通过程序合成了纹理图象。

Study on Texture-image Synthesis Based on BP Algorithm

To solve the present problems in texture-image synthesis,applying the Artificial Neural Network BP algorithm so that the texture- image synthesis is accomplished smoothly. The texture-image analysis and feature extraction is used to obtain images of a variety of parameters, and then using the Artificial Neural Network BP algorithm, a BP network model is established the original image as a function of the characteristics of the BP network input, through the Study-Training to determine the image of the artificial neural network structure, and eventually through the process of synthesis of texture-images.