Effective Disaster Management Using Modified Convolutional Neural Network

The ecological, social, and economic damage caused by fire disasters is largely attributable to human error. Disaster management systems would benefit greatly from early fire detection and autonomous response in order to reduce these losses. Thus, this research work proposes an early-stage fire detection mode by using a Modified Convolutional Neural Networks, which is fine-tuned for CCTV surveillance cameras to identify fire in varient settings, both indoor and outdoor. An adaptive sorting mechanism is used here to ensure the CCTV system's ability to respond autonomously. The experimental results justifies the applicability of the suggested model for enabling efficient fire disaster management and authenticate the enhanced accuracy of fire detection system related to other advanced methods.