System for Screening Objectionable Images Using Daubechies' Wavelets and Color Histograms

This paper describes WIPETM Wavelet Image Pornogra phy Elimination an algorithm capable of classifying an image as ob jectionable or benign The algorithm uses a combination of Daubechies wavelets normalized central moments and color histograms to provide semantically meaningful feature vector matching so that comparisons be tween the query image and images in a pre marked training set can be performed e ciently and e ectively The system is practical for real world applications processing queries at the speed of less than sec onds each including the time to compute the feature vector for the query Besides its exceptional speed it has demonstrated recall over a test set of images found from objectionable news groups It wrongly classi ed of a set of benign images obtained from various sources For di erent application needs the algorithm can be adjusted to show recall while wrongly classifying only of the benign images