Semantic scene understanding using thermal images has received great attention due to the advantage that thermal imaging cameras could see in challenging illumination conditions. However, thermal images are lack of color information and the edges in thermal images are often blurred, making them not very suitable to be directly used by existing semantic segmentation networks that are designed with RGB images. To address this problem, we propose a cross-modal edge-privileged knowledge distillation framework, which utilizes a well-trained RGB-Thermal fusion-based semantic segmentation network with edge-privileged information as the teacher, to guide the training of a semantic segmentation network as the student. The student network only uses thermal images. The experimental results on a public dataset demonstrate that under the guidance of the teacher, the student network achieves superior performance over the state of the arts using only thermal images.

The performance of the zero-effort-miss/zero-effort-velocity feedback guidance algorithm is evaluated through practical space application examples. The zero-effort-miss/zero-effort-velocity feedback guidance algorithm is, in general, not an optimal solution; however, it is an optimal solution in a uniform gravitational environment. It is also conceptually simple and easy to implement and, thus, has great potential for autonomous onboard implementation. It is shown that, for some classic ballistic missile intercept and asteroid intercept scenarios, the zero-effort-miss/zero-effort-velocity algorithm can even compete with corresponding open-loop optimal solutions, while its feedback characteristics make it more suitable to deal with uncertainties and perturbations. By employing the zero-effort-miss/zero-effort-velocity algorithm in the highly nonlinear orbital transfer and raising problems and comparing with corresponding open-loop optimal solutions, its simplicity and near-optimality are further verified.

This paper investigates the optimization approach to generate waypoints for the Mars landing problem in the context of employing the zero-effort-miss/zero-effort-velocity feedback guidance algorithm. For a power-limited engine, the waypoint optimization problem in the presence of state constraints is converted to an equivalent standard quadratic programming problem, which can be solved efficiently. In the case with a thrust-limited engine, by introducing a continuously differentiable function to approximate the standard saturation function, the optimal waypoint can be determined using open-source optimization software. This novel idea exploits parameter optimization techniques for feedback control implementation, thus, it can combine the advantages of open-loop and closed-loop methods to achieve near-optimal performance with acceptable robustness, while meeting various practical constraints and requirements.