Geometric Feature Based Face-Sketch Recognition
This paper presents a novel facial sketch image or face-sketch recognition approach based on facial feature extraction. To recognize a face-sketch, we have concentrated on a set of geometric face features like eyes, nose, eyebrows, lips, etc and their length and width ratio because it is difficult to match photos and sketches because they belong to two different modalities. In this system, first the facial features/components from training images are extracted, then ratios of length, width, and area etc. are calculated and those are stored as feature vectors for individual images. After that the mean feature vectors are computed and subtracted from each feature vector for centering of the feature vectors. In the next phase, feature vector for the incoming probe face-sketch is also computed in similar fashion. Here, K-NN classifier is used to recognize probe face-sketch. It is experimentally verified that the proposed method is robust against faces are in a frontal pose, with normal lighting and neutral expression and have no occlusions. The experiment has been conducted with 80 male and female face images from different face databases. It has useful applications for both law enforcement and digital entertainment.