Fire Detection Using Image Processing in Industries

Nowadays, fire in the workplace causes significant damage because it is not used early on due to a lack of awareness. As a result, all machines in the industry could be damaged. The most prevalent causes of fire include static electricity, overheating, error by humans, external factors, and mechanical failure. This study addresses concerns with existing solutions by incorporating functionality such as alert and detection. To get real-time detection results, the fire detection unit was created using PythonCV and an image processing model. To detect fire images, an image pre-processing technique called the Haar Cascade Classifier was used. The trained model was tested and compared with other cutting-edge methods. A big fire scale and negative photos of fire in various indoor and outdoor environments were used. The video camera has no specific requirements. The camera will keep monitoring the situation in the room as long as the device is powered ON. If the camera detects a fire, the image is evaluated using an image processing model, the buzzer sounds, and an alarm message is delivered to the user's smartphone via Email.