Local contrast methods are efficient approaches for infrared (IR) small target detection. However, since its baseline is obtained locally, a local contrast method can work well only when the target is locally prominent, which is not always the case when the background is very complex. In this paper, a new detection framework named true global contrast measure (TGCM) is proposed, in which its baseline is obtained by global decomposition, so that it can handle situations where the target is not locally salient. For example, when it is adjacent to some bright background. The complex edges / corners and noises that are easily lost during decomposition will be suppressed through some simple but ingenious design: A Gaussian filtering operation is adopted to suppress noises, and a local max dilation is utilized on the separated baseline to maintain as much edge / corner information as possible. The residual information during decomposition will be used as a weighting function to further suppress clutters. Experiments on both real and simulated IR images demonstrated the effectiveness of the proposed method. Compared to some existing algorithms, it has advantages in target enhancement and background suppression. Especially, it can achieve good performance in certain scenarios where the target is not locally prominent.

Semantic segmentation in remote-sensing (RS) images is a critical component for advanced RS scene understanding. Traditional methods learn to segment using direct supervision of paired data, yet fall short by not fully addressing the non-causal biases behind RS images. These include intra-image biases like varying background and object appearance, and inter-image biases such as inconsistent geographical environments and class distributions, which can mislead models to infer by relying on these biases. Given the high cost and effort needed for pixel-wise annotation, the limited volume of RS images exacerbates this problem. In this paper, we propose a novel intra- and inter-image causal intervention (I3-CI) learning paradigm, designed specifically to mitigate the influence of both biases on RS semantic segmentation. Specifically, to reduce the intra-image biases, our I3-CI utilizes intra-image contrastive learning to promote closer alignment of features from the same category and distancing those of disparate categories in the same image. Consequently, the compact feature space is more robust to the varying appearance. As for eliminating the inter-image level biases, our I3-CI further explores the inter-image contrastive learning of all pixels within the entire dataset. To overcome the difficulty of learning from all pixels of the dataset simultaneously, our I3-CI introduces a set of proxy prototype features to keep track of the global centroids for the features of different categories. We substantiate the I3-CI paradigm’s efficacy through rigorous testing on the LoveDA, Vaihingen, and Potsdam datasets with an average improvement of 0.55% on the mIoU metric, proving the value of causal interventions in achieving robust and accurate semantic segmentation.

The tomographic synthetic aperture radar (TomoSAR) technique has been gaining attention because it can retrieve the 3-D structures of urban buildings by synthesizing apertures along the elevation direction. However, most existing TomoSAR methods in literature calculate elevations pixel by pixel and overlook the correlation between elevations, leading to low accuracy and efficiency. To solve these problems, this study introduces an efficient region-based 3-D urban building reconstruction method that incorporates different geometric primitives (i.e., points, planes, and models). Specifically, the proposed method, under the constraints constructed by different geometric primitives, follows three steps to reconstruct the box-like models of urban buildings: 1) it detects double-bounce regions and reconstructs box-like models based on plane sweeping and region growing; 2) it reconstructs box-like models based on multiplane fitting and optimization for nondouble-bounce regions; and 3) it regularizes box-like models based on building layout priors (e.g., collinearity and proximity). The experimental results on two datasets show that the proposed method can efficiently produce reliable results and outperforms several existing methods both qualitatively and quantitatively.