A Rapid Smoke Detection in Real-Time CCTV Videos Using Non-Rigid Bodies Algorithm

World-changing technologies have rolled out in recent years to address an ever-increasing demand for safety and security. While fire accidents are increasing due to a variety of factors, the need for enhanced technology to detect smoke and fire at an early stage has been declared. An algorithm for detecting incipient stage smoke under any backdrop conditions has been proposed in this study using non-rigid regional bodies. The feature vector for multi-background frames was optimized in order to identify the swift and cost-effective way. Consequently, the non-rigid bodies algorithm (NRBA) was created to detect the onset of smoke or fire with a detection accuracy of 1 second for the high frame rate camera and 4 seconds for the low frame rate camera.