Face spoof detection using image distortion analysis and image quality assessment
Shilpa Unnikrishnan, Ansiya Eshack · 2016
Secure face spoof detection systems demand the capability to identify whether a face is from a real client or a portrait from a spoofer. Spoofing induces distortion in the image and also degrades the image quality. Analysis of distortion and the quality assessment of an image to identify spoof attack is the main consideration here. The existing methods in image distortion analysis, extracts the features that capture the facial details. It extracts four different features (specular reflection, blurriness, chromatic moment, and color diversity) to form the IDA (Image Distortion Analysis) feature vector. The existing methods in image quality assessment, extracts several general image quality features to form IQA (Image Quality Assessment) feature vector. The designed system utilizes a hybrid scheme of both IDA and IQA. In addition, it also extracts the Fourier based and Wavelet based features of the image. A Neural Network (NN) classifier is used for the training. It is seen that the designed hybrid system face spoof detection achieves high performance than the existing system.