A novel fusion method for intense heat detection and localization solely based on surface temperature with multi-mode DEA feedback

Heat detection and localization are vastly being used in various fields such as navigation systems, autonomous systems in industry, medical and research. To provide a more efficient detection and localization system, feedback capability is significant. Towards that goal, previously shown simultaneous multi-mode Dielectric Elastomer Actuator (DEA) technology can be used together with MEMS technology in a combined manner for development of an interactive sensory system with auditory and haptic feedback. By driving a certain DEA topology with low-frequency and additional high-frequency signal components it is possible to produce multiple independent outputs, like linear movement and sound generation to develop interfaces with simultaneous but independent acoustic and tactile feedback capability. In this study, a single IR sensor is introduced to detect heat in indoor environments and the pixels data from the sensor is transformed into a 32x32 matrix for heatmap generation. The study applies array manipulation to localize and calculate an approximate distance of the target from the source based on surface temperature. A YOLOV8 model is used with a custom dataset for objects and human detection based on generated thermal images derived from our function. The integrated DEA control can adjust feedback autonomously based on the sensor field of view in real-time. The study shows an overall accuracy of more than 90% in continuous based distance estimation when evaluated with depth tracking data from a stereo camera. For detection and classification of heated objects (human, electronics, hot objects) while navigating, the accuracy is above 95%.