Height Estimation

Height estimation, the process of determining the vertical dimension of objects or terrain, is crucial across diverse fields, from forestry and urban planning to robotics and autonomous driving. Current research focuses on improving accuracy and efficiency using various techniques, including deep learning models like convolutional neural networks (CNNs) and vision transformers (ViTs), often incorporating multi-modal data fusion (e.g., combining SAR, LiDAR, and optical imagery) or leveraging innovative approaches such as classification-regression paradigms and belief-space control for active tactile perception. These advancements enable more precise measurements for applications ranging from carbon stock assessment and 3D city modeling to improved robotic manipulation and safer autonomous navigation.