Accurate inversion of low altitude atmospheric temperature and humidity is crucial for weather forecasting and climate monitoring. This letter introduces the MR-TH method, a deep learning approach that uses convolutional neural networks and Transformer architecture to invert low altitude three-dimensional atmospheric temperature and humidity distribution from airborne microwave radiometer data. By capturing nonlinear relationships and spatial correlations, MR-TH improves the inversion accuracy of traditional methods. This network is trained and validated using onboard flight data, reanalysis products, and radiosonde measurements. The results indicate that the mean square error (MSE) of temperature inversion for MR-TH is 0.3-1.5 K and the humidity MSE is 0.2-2.0 g/kg, with an accuracy improvement of over 15% compared to the BP neural network method within the range of 1-5 km altitude. MR-TH also shows a high correlation (>90%) with radiosonde data. MR-TH provides a feasible solution for improving the accuracy of atmospheric parameter inversion from airborne microwave radiometer observation data.

The Aperture Synthetic Radiometer (ASR) system consists of an antenna array along with subsequent receiving channels and data processing modules, and it has been widely applied in Earth observation fields. The arrangement of the antenna array directly affects the system’s sampling coverage in the spatial frequency, which in turn determines its imaging quality. Therefore, we propose a cooperative optimization method for antenna arrays based on deep reinforcement learning, aiming to improve the sampling coverage of the ASR system, thereby enhancing its imaging performance. Experimental results show that the final stable coverage value of the proposed method is 10.545%, representing a 41.2% improvement over the lowest value. Furthermore, it outperforms other comparative methods in both coverage performance and reconstructed image quality, demonstrating its effectiveness in optimizing antenna array layout and enhancing the imaging quality of the system.

In 1-D mirrored aperture synthesis (MAS), the antenna array arrangement and metal reflector are crucial in determining the rank of the transformation matrix. Accurate cosine visibility is achievable only when the transformation matrix is full rank. However, the anti-phase characteristic of the metal reflector introduces non-zero elements of “-1” into the matrix, leading to rank deficiency. This letter proposes a method of using artificial magnetic conductor (AMC) with in-phase reflection property instead of metal reflector to ensure that the transformation matrix only contains non-zero elements “1”. Based on this property, the rank of both linear and nonlinear arrays is verified. The results indicate that AMC can effectively enhance the rank of the transformation matrix, potentially achieving full rank. Additionally, further verification is performed on the reconstruction of trapezoidal extended source scene using two types of arrays. The results demonstrate that AMC-based 1-D MAS can achieve a low root-mean-square error (RMSE), significantly improving the quality of the reconstructed images.

Accurately measuring and inverting air parameters, such as air temperature and humidity, is crucial for weather forecasting, climate research, and environmental monitoring. In this letter, we propose an inversion method based on the transformer model to accurately estimate the spatial distribution of air temperature and humidity. Compared with traditional methods, the transformer model demonstrates superior ability in capturing nonlinear relationships and spatial dependencies in observational data, thereby improving inversion accuracy. Experiments conducted on real observational data have shown that compared to traditional techniques, the proposed method achieves a reduction of over 4.8% in the root mean square error (RMSE) of air temperature and over 14.2% in humidity estimation, demonstrating its high accuracy and reliability in inverting air temperature and humidity. This method provides a new approach for advancing air parameter inversion technology.

Current 3D detection methods, whether single-modal or multimodal, face notable limitations. Single-modal detectors, using either camera or LiDAR, struggle with spatial accuracy and object differentiation due to insufficient depth information or difficulty distinguishing semantically similar objects. Existing multimodal fusion techniques, while improving performance, often suffer from high computational costs, false positives, and complex architectures, especially when utilizing anchor-based pipelines. To address these challenges, we propose an efficient pointwise fusion method that directly extracts point features from enhanced RGB images and fuses them with corresponding point cloud features, preserving essential spatial and semantic information. This fused data is then processed through a three-dimensional neural network, significantly improving inference speed and detection performance. Our framework is designed for multi-class 3D object detection, leveraging the complementary strengths of LiDAR and camera data without the need for multiple backbones or complex synchronization steps. Extensive experiments on the KITTI benchmark demonstrate that the proposed method outperforms state-of-the-art LiDAR-camera fusion techniques, achieving 92.5% AP for 3D detection and 95.41% AP for BEV detection, making it particularly suitable for autonomous driving systems. These results highlight the effectiveness of the proposed fusion strategy in balancing accuracy, computational efficiency, and robustness in complex 3D environments.

For the tilted mirrored aperture synthesis (MAS), it can obtain different fields of view (FOV) by adjusting the inclination angle of the reflector. However, the tilted MAS microwave radiometer system is more complex. The antennas need to receive both direct signals and signals reflected by different reflectors, and reflector errors will affect the performance of the tilted MAS system. At present, there is no effective method to correct various errors in tilted MAS systems (especially reflector errors), which results in poor quality of brightness temperature (BT) image reconstruction for the tilted MAS. In this paper, we design a network to reconstruct BT images based on CNN and Transformer, and present a method to reconstruct the tilted MAS BT image by using this network. Based on the analysis of the FOV and spatial resolution of the tilted MAS system, a method for collecting the measured training dataset is presented. Using the measured training dataset for network training, the network learns the information of different errors in the tilted MAS system, including the error information of different reflectors and the error information of amplitude and phase. The simulation and experimental results show that the proposed method can obtain BT images with higher reconstruction quality than the impulse matrix reconstruction method.

To increase the spatial resolution of passive microwave radiometry, mirrored aperture synthesis (MAS) was presented. In this letter, the method of cosine visibility extension (CVE) is proposed to further enhance the spatial resolution of 1-D MAS. In the CVE method, a convolutional neural network (CNN) is used to learn the distribution of the cosine visibility (CV), specifically the relationship between the low- and high-frequency CV distributions of various scenes. Then, the high-frequency CVs are estimated by the CNN according to the low-frequency CVs obtained by MAS. The high- and low-frequency CVs are combined in MAS image reconstruction to enhance the spatial resolution of MAS. The simulation and experiment indicate that the CVE method can effectively enhance the spatial resolution of MAS.

For 1-D aperture synthesis (1-D AS) radiometers, truncated sampling occurs in the frequency domain due to the system baseline limitation. Therefore, there is an obvious Gibbs oscillation in the reconstructed image. To solve this problem, an imaging method based on a 1-D convolutional neural network (1-D CNN) is proposed in this article. Compared with deep learning methods based on 2-D convolutions, the 1-D convolution not only reduces the amount of computation but also produces further performance improvements. The input data of the network are the 1-D visibility function samples, and the output data are the 1-D brightness temperature (BT) samples. The network learns the mapping relationship from the training of the 1-D visibility function samples and 1-D BT samples to complete 1-D AS imaging without any prior knowledge. To verify the performance of this imaging method, simulations and experiments based on the airborne C-band 1-D microwave interferometric radiometer (ACMIR) system are implemented. The simulation and experimental results demonstrate that the proposed AS-CNN method achieves higher performance than the inverse fast Fourier transform (IFFT) method in terms of image quality and Gibbs phenomenon suppression. In the case of an antenna failure and missing baseline, the AS-CNN method proposed in this article can still obtain a BT image with high imaging quality, which shows that the robustness of the network is better than that of the IFFT method.

In passive microwave remote sensing of the Earth, compared with real aperture radiometer, synthetic aperture radiometer (ASR) is a very powerful instrument with many advantages. However, the system design is more complex than the real aperture radiometer, and the hardware ideality is often not guaranteed. The nonideal characteristics of the system hardware will bring a variety of errors to the system, which will cause the Fourier transform relationship between the visibility function and the brightness temperature image to no longer be established, thus reducing the quality of microwave brightness temperature image reconstruction by traditional methods. In this article, a new microwave brightness temperature image reconstruction method for ASR by transformer is proposed. This method uses a specially designed transformer structure to extract the spectrum features in the visibility function. This method learns the mapping relationship between the visibility function and the original scene brightness temperature image through the supervised learning method, and learns as much as possible the spectrum information contained in the original scene brightness temperature image. Moreover, when there are missing baselines, this method will supplement the missing observation frequency information, so as to obtain better reconstructed image quality. With the above-mentioned advantages, this method can suppress the Gibbs oscillation, and greatly reduce the sidelobe. Compared with the existing reconstruction methods, whether missing baselines or not, the proposed image reconstruction method by transformer has advantages in image quality. We verify the performance of this brightness temperature image reconstruction method through simulation and experiment.

In this paper, Mirrored Aperture Synthesis with tilting reflectors (MAS-T) is proposed to improve spatial resolution with fewer antennas for passive microwave remote sensing. The principle of MAS-T is given. The initial experimental results demonstrate that MAS-T can achieve higher spatial resolution with the same antenna array compared with conventional aperture synthesis.

The antenna array arrangement determines the rank of the transformation matrix in 1-D mirrored aperture synthesis (1-D MAS). Only when the transformation matrix is of full rank can cosine visibilities be precisely solved from the transformation equations. However, the existing literature states that full rank cannot be realized with the default perpendicular polarization of antennas. In this letter, the transformation matrices of arrays with various polarizations of antennas are discussed, and a type of full array with a full-rank transformation matrix is found. Based on these results, the existence conditions for an array with a full-rank transformation matrix are obtained. In addition, a search algorithm with a shorter search time than the existing algorithm is designed to find low-redundancy arrays with a full-rank transformation matrix. The simulations verify that cosine visibilities are precisely solved from the transformation equations for the array with a full-rank transformation matrix.

The demand for high-precision sea surface temperature (SST) has been growing rapidly in recent years because SST is one of the key parameters to describe the thermal state of the sea surface. This article analyzes the differences between microwave remote sensing and infrared remote sensing for SST, including the spatial resolution difference and the penetration depth difference. In order to improve the accuracy of retrieved SST, this article proposes a joint inversion algorithm for SST from combining microwave brightness temperature (BT) and infrared BT, which has also taken the influence of wind speed and atmosphere into consideration. Experiments confirm that SST data obtained from the joint inversion algorithm are more accurate than those obtained from the existing inversion algorithms.

Setting two reflectors perpendicular to each other and an array produces mirrored aperture synthesis (MAS), which obtains a higher spatial resolution than aperture synthesis (AS) for the same antenna array. The proposal of the MAS theory and the experimental verification suggest that the spatial resolution can be improved by setting the reflector. This paper proposes a method in which four tilted planar reflectors are set around an antenna array. This method is named two-dimensional (2-D) mirrored aperture synthesis with four tilted planar reflectors (2-D MAS-T). Because 2-D MAS-T uses more reflectors than MAS, the spatial resolution is further improved. The principle of 2-D MAS-T with an antenna array is given in this paper. The relationship between the size of the reflectors, the spatial resolution and the field of view (FOV) is obtained. The simulation and experimental results demonstrate that 2-D MAS-T can achieve a higher spatial resolution with the same antenna array than 2-D AS and 2-D MAS.

In mirrored aperture synthesis (MAS), the existing brightness temperature image reconstruction methods include inverse cosine transform and impulse matrix reconstruction methods. However, the quality of the MAS brightness temperature images reconstructed by the existing methods is still poor and needs to be improved. This article proposes a method of MAS brightness temperature image reconstruction with deep convolutional neural network (CNN). The network includes two fully connected (FC) layers, multiple convolutional layers, and deconvolutional layers, which realize the image reconstruction for MAS. This method uses deep CNN to learn the MAS image reconstruction mapping and system errors, so as to improve the performance of the brightness temperature image reconstruction. Both simulation and experimental results verify that the performance of the proposed MAS-CNN method is better than the existing MAS image reconstruction methods.