一个基于内容的多层次敏感图像过滤框架

A Framework for Content-based Hierarchical Sensitive Image Filtering

互联网中存在大量的敏感信息,特别是敏感图像信息,其毒害远远甚于敏感文本信息。本文探索性地给出了一个基于内容的多层次敏感图像过滤框架,该框架结合黑白名单、图像文本内容过滤、图像内容过滤及相关反馈技术,并采取分流并行方法来解决图像过滤普遍存在的速度问题。

As the harm of image information is far more than that of text information, to solve the problem of much sensitive information appearing on Internet, a hierarchical sensitive image framework is presented, which combines black - white list, image text filtering, image content filtering and correlation feedback technology, and adopts distributed parallel mechanism to solve the speed problem of image filtering.