Thermal infrared (IR) images represent the heat patterns emitted from hotobject and they do not consider the energies reflected from an object. Objectsliving or non-living emit different amounts of IR energy according to theirbody temperature and characteristics. Humans are homoeothermic and hencecapable of maintaining constant temperature under different surroundingtemperature. Face recognition from thermal (IR) images should focus on changesof temperature on facial blood vessels. These temperature changes can beregarded as texture features of images and wavelet transform is a very goodtool to analyze multi-scale and multi-directional texture. Wavelet transform isalso used for image dimensionality reduction, by removing redundancies andpreserving original features of the image. The sizes of the facial images arenormally large. So, the wavelet transform is used before image similarity ismeasured. Therefore this paper describes an efficient approach of human facerecognition based on wavelet transform from thermal IR images. The systemconsists of three steps. At the very first step, human thermal IR face image ispreprocessed and the face region is only cropped from the entire image.Secondly, Haar wavelet is used to extract low frequency band from the croppedface region. Lastly, the image classification between the training images andthe test images is done, which is based on low-frequency components. Theproposed approach is tested on a number of human thermal infrared face imagescreated at our own laboratory and Terravic Facial IR Database. Experimentalresults indicated that the thermal infra red face images can be recognized bythe proposed system effectively. The maximum success of 95% recognition hasbeen achieved.
展开▼
