Independent monitoring as well as video surveillance have a lengthy history. In both controlled indoor and outdoor environments, many currently available devices can accurately monitor human mobility. As a constant part of our everyday lives, low-light conditions have a significant impact. Nevertheless, one of the biggest challenges in visual surveillance is still object detection at night. There has been a rise in poor light image studies, especially in the area of image improvement, but no relevant database serves as a standard. One use of object detection is the remote or centralized management of a large number of security and video surveillance devices. It is suggested that night vision monitoring could benefit from the use of an object detection technique. The method relies on detecting motion. PIR sensors might pick up on unnoticed motion to kick off the search. Due to motion prediction, this method works well in practice for night-time detection. Furthermore, we discuss our interesting and insightful findings concerning the impacts of low light on the object detection job on developing a Deep Learning (DL) method. If an object is spotted, an alert message is sent to the user's registered mobile phone through GSM technology, and send an email with a half-minute video clip of the surroundings. Our investigation on dark images is meant to pave the way for more studies in the low-light domain.01 © 2023 Elsevier B.V., All rights reserved.