<p>Farm fire has become one of the major concerns in the agriculture economy. The sensor-based models and computer vision-based models are introduced in the contemporary literature for the detection of fire. However, these methods are ineffective in detecting and exhausting the fire in real time due to less smoke, different environmental condition and also suffer from low accuracy and high false alarm rate. This problem can be effectively solved if the fire is detected in the real time and correspondingly the exhaustion can be performed automatically at the same time. Therefore, in this paper, the smart farm management system (SFMS) using Internet of Things (IoT) and Deep Convolutional Neural network is proposed which took the benefit of IoT network model for monitoring the farm area and the sensed data is fed to MobinetV2 model for the detection of any fire incidence. In the case of fire, IoT network automatically activates the water sprinkler for the real time exhaustion of fire. Further, the web interface is linked with the SFMS for the notification, alerts about the status of farm. The simulation results have shown that the proposed model is robust in detecting fire or no fire with the accuracy of 99.9%, 99.82% and 99.7% in Green Farm with no fire, Fire in progress and Fire in the cloudy environment. The proposed model performed comparatively better than state of art approaches and effective in exhausting the fire.</p>