基于显著性信息的数字绘画图像自主分类系统

Digital painting image autonomous classification system based on saliency information

数字绘画图像中存在大量内容增加了自主分类的难度,因此文中研究一种基于显著性信息的数字绘画图像自主分类系统。该系统由图像预处理模块、显著性信息特征提取模块以及卷积神经网络自主分类模块三部分组成。图像预处理模块将输入数字绘画图像通过中值滤波方法过滤,然后发送至显著性信息特征提取模块,显著性信息特征提取模块接收过滤后的图像,利用流形排序算法计算图像显著性信息获取得到显著性信息图,将显著性信息图输入卷积神经网络中,利用卷积神经网络分析输入样本建立分类模型,将待分类图像输入卷积神经网络中,利用已训练分类模型实现数字绘画图像自主分类,并将分类结果发送至用户界面。实验结果表明,该系统数字绘画图像分类精度高达99%以上。

There are a large number of content in digital painting images,which increase the difficulty of autonomous classification.A digital painting image autonomous classification system based on saliency information is researched.The system is imposed of the image preprocessing module,saliency information feature extraction module,and convolutional neural network autonomous classification module.The image preprocessing module is used to filter the inputted digital painting image by means of the median filtering method and send it to the saliency information feature extraction module.The saliency information feature extraction module is used to receive the filtered image and calculate the saliency information by means of the manifold ranking algorithm to obtain the saliency information map.The saliency information map is inputted into the convolutional neural network,and the sample is analyzed by means of the convolutional neural network to establish the classification model.The image to be classified is inputted into the convolutional neural network to realize the autonomous classification of the digital painting image through the trained classification model,and the classification result is sent to the user interface.The experimental results show that the classification accuracy of this system for digital painting images is more than 99%.