Abstract:
From the research project of digital technology communication for agriculture field fire by unmanned aerial vehicle by image processing and computer Vision. It was tested on the ground according to the real situations. The images and video getting from the process were analyzed and evaluated to develop as a tool a forest fires to look for the ways of warnings and prevention. The results of the operation were able to create three custom four-propeller aircraft, two 500-meter transmission range, and compared with one DJI finished drone model. It was taught in 1) tensorflow object detection by raspberry pi, 2) Yolo object detection by raspberry pi, and 3) object detection by Yolo flame through Matlab.
And all techniques are developed in conjunction with shadow or gray detection or color detection techniques, image subtraction and defect detection technique, tracking technique and video background subtraction technique. In the next step, the image processing and visualization techniques through the computer screen will be brought to the notification via the application on the smartphone