Face detection using quantized skin color regions merging and wavelet packet analysis

Abstract

Detecting and recognizing human faces automatically in digital images strongly enhance content-based video indexing systems. In this paper, a novel scheme for human faces detection in color images under nonconstrained scene conditions, such as the presence of a complex background and uncontrolled illumination, is presented. Color clustering and filtering using approximations of the YCbCr and HSV skin color subspaces are applied on the original image, providing quantized skin color regions. A merging stage is then iteratively performed on the set of homogeneous skin color regions in the color quantized image, in order to provide a set of potential face areas. Constraints related to shape and size of faces are applied, and face intensity texture is analyzed by performing a wavelet packet decomposition on each face area candidate in order to detect human faces. The wavelet coefficients of the band filtered images characterize the face texture and a set of simple statistical deviations is extracted in order to form compact and meaningful feature vectors. Then, an efficient and reliable probabilistic metric derived from the Bhattacharrya distance is used in order to classify the extracted feature vectors into face or nonface areas, using some prototype face area vectors, acquired in a previous training stage.

Tracking people under heavy occlusions by layered data association

“…Skin colors are distributed differently from hair colors, spreading more continuously, yet within a bounded range [5]. As we found that the illumination conditions have a greater influence on detection of skin than they do on hair, we decided to model the skin colors based on three illumination categories, namely, bright, standard, and dark.…”

“…Weights in (5) are chosen by a search over interval [0,1] in steps of 0.1 units so as to maximize correct hair pixel detection over a training set.…”

Skin Segmentation Using GMM Classifier and Texture Feature Extraction

“…In six algorithms, The YcbCr approach observes skin color as 2D model and other as 3D model, they using Genetic Algorithm (GA) to training the boundary of several color models, their experiment shown that YcbCr color space is more recall oriented and RGB is a good tradeoff between recall and precision [6]. Christophe and Georgios using vector quantization technique [12], to quantize the image color, and construct a 3D skin color model that distributed in HSVand YcbCr color space, both colo models is constructed by 950 skin color samples which have been extracted from various still images and video frames. Jones and Rehg apply visualization techniques [7] to fetch out the shape of skin color distribution.…”

Indoor object recognition through human interaction using wavelet features

“…In [2,3], mean values and the corresponding variances were computed from face images to characterize different faces. Specifically, 3 mean values and 3 variances of the approximation image and 15 variances of details images form the feature vector.…”