ABSTRACT Generally, there are two problems restrict the intracranial haemorrhage (ICH) segmentation task: scarcity of labelled data, and poor accuracy of ICH segmentation. To address these two issues, we propose a semi‐supervised ICH segmentation model and a dedicated ICH segmentation backbone network. Our approach aims at leveraging semi‐supervised modelling so as to alleviate the challenge of limited labelled data availability, while the dedicated ICH segmentation backbone network further enhances the segmentation precision. An augmented multiple perturbation dual mean teacher model is designed. Based on it, the prediction accuracy may be improved by a more stringent confidence‐weighted cross‐entropy (CW‐CE) loss, and the feature perturbation may be increased using adversarial feature perturbation for the purpose of improving the generalization ability and efficiency of consistent learning. In the ICH segmentation backbone network, we promote the segmentation accuracy by extracting both local and global features of ICH and fusing them in depth. We also fuse the features with rich details from the upper encoder during the up‐sampling process to reduce the loss of feature information. Experiments on our private dataset ICHDS, and the public dataset IN22SD demonstrate that our model outperforms current state‐of‐the‐art ICH segmentation models, achieving a maximum improvement of over 10% in Dice and exhibiting the best overall performance.

High-precision localization of the Mars rovers is fundamental for path planning and safe navigation toward exploration targets during Mars missions. In cross-site visual localization, image matching is the key step to obtain corresponding points connecting images from different sites. The cross-site visual localization method based on Affine SIFT (ASIFT) is used in Tianwen-1 mission but is constrained in regions of Mars with poor texture and large viewpoint invariance. In this article, we propose a cross-site visual localization methodology of Mars rover based on self-supervised keypoint extraction and robust matching. The self-supervised keypoint extraction network, which is called MRSS-Net, uses multiscale deformable structures (MSDSs) during the feature encoding stage to enhance the network’s ability of extracting invariant features in regions with large viewpoint variations and improve the rate of identical points for cross-site images with poor texture. In addition, we develop self-attention descriptor enhancement mechanism (SADEM) to distinguish local features in repetitive patterns. The robust matching, which is called adaptive 2-D–3-D matching, uses GNC dead-reckoning (3-D priori information) to construct the initial coarse matching domain and homography matrix (2-D information) to construct a progressively shrinking refined matching domain. We compared our method against ASIFT based cross-site visual localization model and advanced deep learning algorithms and evaluate the performance using NaTeCam images collected during the traversal of four long-distance traversals (a total of 44 Martian sol sites) by Zhurong rover. The experimental results show that our framework reduces the localization error by 12.5% and improves localization robustness by 50.8%, compared with ASIFT-based cross-site visual localization method used in Zhurong rover. In addition, our method outperforms state-of-the-art deep learning techniques and ensures the current accuracy of cross-site visual localization for Mars rover, while significantly increasing the level of automation.