Visual-based robot pose estimation is a fundamental challenge, involving the determination of the camera’s pose with respect to a robot. Conventional methods for camera-to-robot pose calibration rely on fiducial markers to establish keypoint correspondences. However, these approaches exhibit significant variability in accuracy and robustness, particularly in 2D keypoint detection. In this work, we present an end-to-end pose estimation approach that achieves camera-to-robot calibration using monocular images and keypoint information. Our method employs a two-level nested U-shaped architecture, featuring a bottom-level residual U-block to extract richer contextual information from diverse receptive fields to enhance keypoint refinement. By incorporating the perspective-n-point (PnP) algorithm and leveraging 3D robot joint keypoints, we establish correspondence of 3D coordinate points between the robot’s coordinate system and the camera’s coordinate system, facilitating accurate pose estimation. Experimental evaluations encompass real-world and synthetic datasets, demonstrating competitive results across three distinct robot manipulators.

The Cosmic-ray Soil Moisture Observing System is a promising soil moisture measurement network. It can measure soil moisture at an intermediate spatial scale with a single sensor. In this letter, the measured cosmic-ray neutron counts during the Heihe Watershed Allied Telemetry Experimental Research were used to evaluate the capabilities of the cosmic-ray probe in soil moisture retrieval at a heterogeneous farmland. The Cosmic-ray Soil Moisture Interaction Code model was utilized to model the interaction between the measured neutron counts and the soil water content. Soil moisture at the footprint scale of the cosmic-ray probe obtained using a wireless sensor network (SoilNET) was used as the calibration and validation data. The results show that the cosmic-ray probe is capable of monitoring the hourly heterogeneous soil moisture dynamics at the intermediate spatial scale in a noninvasive way. Moreover, the informative measurement depth of the cosmic-ray probe can also be derived and is consistent with the soil moisture results.

A multi-sources data assimilation system has been developed for the catchment scale land water and energy cycle researches. The land surface model, microwave radiative transfer model and ensemble Kalman filter have been coupled in this system, the high performance computing was also considered. This system is being used in the data assimilation of soil moisture, soil temperature, microwave brightness temperature and snow water equivalent at catchment scale.

MODIS data and in situ data were used to establish a real thermal inertia model to retrieve soil thermal inertia of Heihe River Basin (HRB) and a soil thermal inertia-soil moisture model to calculate the soil moisture of the basin. Meanwhile, the data of two observation stations in the basin was used to validate the result of retrieval. The correlation coefficients of those between retrieved and observed at two stations were 0.853 and 0.917, respectively. The result shows the method meet the requirement of soil moisture inversion.

A preliminary study has been conducted to test the effectiveness of the stepwise canonical discriminant approach to making spatial discrimination of salt-affected soil surfaces, using the Landsat TM data taken in the middle reaches of the Heihe River, China. We regard a saline area as being composed of salt-affected patches differ in degree and try to make discrimination of salt-affected soil surfaces by stepwise canonical discriminant approach. Five indices (B4, Wetness, SI, NDSI, Fe2O3) were selected as discriminating variables, and two canonical discriminant functions (y1, y2) were determined for the three groups using these five discriminating variables. When group centroids and individual cases were plotted on the plane determined by two canonical discriminant functions y1and y2, the groups were quite distinct and their centroids were well separated. These results suggest that the stepwise canonical discriminant approach is promising for spatial discrimination of salt-affected soil surfaces.

Radar remote sensing has shown its potential for retrieving soil moisture from soil surfaces. Since the backscattering process is also influenced by soil roughness, the characterization of this roughness is crucial for an accurate estimation. The algorithm proposed in this investigation aiming to obtain the roughness parameters for every SAR pixel which could facilitate the derivation of soil moisture in virtue of multi-angle ASAR images, and combined with a semi-empirical calibration for the correlation length. An application of the means was performed in the middle reaches of the Heihe river basin and achieved satisfactory results (RMSE less than 6 vol %).

Incorporating the spatial correlation information into the land data assimilation system provides the opportunities to improve the analysis quality of the assimilation system. In order to make use of the spatial autocorrelation information effectively in the land data assimilation system, we used the geostatistics to explore and describe the spatial variations of the microwave brightness temperature data. The corresponding experimental semivariogram model, covariance model, nugget effect, sill value and range were calculated. Then these results were applied to the localization of the microwave remote sensing observation error covariance successfully, and played an important role regarding the enhanced performance of the direct radiance assimilation system.