<![CDATA[Fire events are disasters that occur because of the expansion of the source of the fire that propagates to potentially burned objects. In addition, these events are also often caused by irresponsible human negligence. Therefore risk management is carried out in anticipation of minimizing the impact of losses. One of them is by using a quadcopter to monitor and detect objects. This study was designed to detect hotspot objects in a room through image processing using the ArDrone front camera for sending raw frame data images at UDP port 5555. Image processing is implemented on the ROS platform which is used to connect quadcopter with GCS using OpenCv. This system utilizes the Hue Saturation Value (HSV) Color Filtering and YCbCr Color Space methods. The results of this study are in the form of a system on ArDrone that is capable of detecting hotspot objects. Testing on the system is done by doing detection tests with different distances with reference to the position of the hotspot object, testing the range of HSV value parameters along with YCbCr values in each condition, testing with a certain speed by reference to the position of hotspot objects and testing the system response performance. The results obtained are effective detection distance is 8 meters with an overall percentage of 83.33%, the speed of the appropriate ArDrone quadcopter if detected is with a speed of 1m/s and the system response performance to detect hotspot objects by 0.022 seconds.]]>
Copyrights © 2019
