Recognition and Evaluation of Facial Expression and Emotion of Students Using Surveillance Cameras with Thermal Detectors

SURVEILLANCE CAMERAS are one among the most extensively used technologies for corporeal security. In this paper, we propose an approach to figure out the bodily emotions and facial expressions of students while conducting an examination. Generally students tend to commit malpractices during exams. Irrespective of its consequences. Here we introduce hidden surveillance cameras with inbuilt thermal imaging detectors to identify the culprits by their emotions. A video collection device is known as surveillance cameras. This is positioned at specific locations in the school or colleges where main examinations are conducted. As surveillance cameras accomplishments have become intensified most contemporarily. The technology has been evolved in such a way that, it strives to accomplish automatically by operating via facial expression and emotion recognition using the information in the form of images obtained from a surveillance system with thermal detectors. In this paper, we extend this approach for the recognition of facial expression, using test images. This can either be colour images or gray images. The emotions are retrieved using thermal imaging detectors. Here a range is set for detection of emotions like fear, tension, anger, worried, etc. Once the body emits a certain amount of heat for the range specified, then it is easier to focus on the culprits. For the facial feature vectors there is a generation from the main key point descriptors using Speeded-Up Robust Features. Here every point in the facial feature vector is being brought to a standard condition and succeeding of the probability density function elucidation is engendered. The interspace between two probability density function elucidators are calculated using Kullback Leibler divergence. Here utilisation of Mathematical equation is performed for selecting indubitable enactment probability density function elucidators for each lattice, that are used for inceptive organizing. Eventually, besides this, we are using a facial expression classifier known as the SVM (Support Vector Machine) classifier to distinguish among various expressions. The output is shown as the recognition result. The proposed approach shows exceptional performance when applied for the Test Facial Expression database. However, the surveillance cameras are also exploited for various purposes. Also these technologies have chances of being utilized malevolently. The privacy for these type of technology could be seriously contravened. Keywords—Surveillance cameras, Thermal detectors, Support Vector Machine, Speeded-Up Robust Feature, KullBack