Face and Human Detection in Low Light for Surveillance Purposes

Surveillance based on computer vision is the need of the current era. Most of the existing surveillance cameras which are being used by the Security Services and Military Forces works on the sensor-based analysis of conditions, it activates a signal indicating the presence of some movable creature and cannot accurately differentiate between a hare, a deer or a human. The current surveillance methods show high accuracy in presence of daylight but fail miserably in low light conditions because these sensors equipped with normal cameras are not able to capture images with the same accuracy in such conditions. Hence to overcome this problem, we propose a network that is a combination of Human and Face Detection systems to distinguish between different entities and enhance the accuracy of the results. This model works well in extreme conditions such as low-light, blurred images, fog, haze or when most of the body or face is covered.