【摘要】 随着互联网络和多媒体技术的发展,色情图像的传播越发猖獗.为了有效地防止色情图像的传播,本文中提出一种综合利用颜色和纹理特征的色情图像过滤方法.该方法根据色情图像的颜色和纹理特征,建立肤色模型和皮肤模型对图像进行检测识别.利用该方法对752幅色情图像和4000幅正常图像进行过滤实验,结果表明该方法具有较高的准确率.更多还原

【Abstract】 With the development of Internet and multimedia technology, the spread of nude images becomes more and more rampant. In order to solve the problem, a color and texture-based method for nude image filtration is proposed in this paper. In this method, a color model and a texture model are set up according to the characters of the color and texture of images, which are then used to detect nude images. The proposed method is finally applied to the filtration of 752 nude images and 4000 benign images. And the results verify the accuracy of the method.更多还原