Pornographic Images Detection Based on CBIR and Skin Analysis

Abstract

A novel two-stage scheme of pornographic image detection is proposed in this paper. Specifically, we first apply the content-based image retrieval technique to find out whether human are present in the images. Then a detailed skin color analysis is performed to affirm the presence of pornographic content in the images. Experimental results show that the proposed algorithm performs well and fast in detecting pornographic images.

Pornographic image detection utilizing deep convolutional neural networks

“…Furthermore, the appropriate low level features are hard to be designed. In [5], content-based image retrieval technique was applied to find out whether human were present in the images, then 85 a skin color analysis was performed to detect skin region, based on which, the presence of pornographic content in the images was affirmed. Unfortunately, detecting whether a human appeared in an image was very difficult.…”

“…Unfortunately, many illegal websites post a lot of pornographic images which may cause serious social problems. For instance, pornographic images can be seen by underage children or other non- 5 authorized institutions. Therefore, it is necessary for us to design an effective pornographic images detection method.…”

“…Approach based on textual cues has been utilized [2], it searched keywords near the web images to find pornographic contents. 15 Recent developments in computer vision have motivated many image contentbased methods for pornographic content detection (could be roughly divided into rule-based methods [3], image retrieval-based methods [4] [5] and learning-based methods [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]). These methods are based on the rich visual information extracted from the original images (e.g.…”

Nude Detection in Video Using Bag-of-Visual-Features

“…In [23], the first step of the system proposed is to use content-based image retrieval to determine whether the image contains humans in it. This retrieval step is based on color and shape features.…”

Research on key thechnologies of pornographic image/video recognition in compressed domain

“…The skin-driven features, such as area, shape and color of skin regions, have been adopted widely in pornographic image/video recognition in pixel domain [4][5][6] . In the proposed compressed domain based method, skin color also takes the most important place.…”