Search

Pulse pile-up presents a significant challenge in nuclear radiation measurements, particularly in neutron-gamma pulse shape discrimination, as it causes pulse distortion and diminishes identification accuracy. To address this, we propose an optimized Support Vector Regression (SVR) algorithm for correcting pulse pile-up. Initially, the Dung Beetle Optimizer (DBO) and Whale Optimization Algorithm (WOA) are integrated to refine the correction process, with performance evaluated using charge comparison methods (CCM) for pulse shape discrimination. Leveraging prior knowledge from simulated data, we further analyze the relationships between various types of pulse pile-ups, including their combinations, inter-peak distances, and the accuracy of corrections. Extensive experiments conducted in a mixed neutron-gamma radiation field using plastic scintillators demonstrate that the proposed method effectively corrects pulse pile-up and accurately discriminates between neutron and gamma. Moreover, our approach significantly improves the fidelity of pulse shape discrimination and enhances the overall reliability of radiation detection systems in high-interference environments.

Remote sensing object detection (RSOD) plays a crucial role in resource utilization, geological disaster risk assessment and urban planning. Deep learning-based object-detection algorithms have proven effective in remote sensing image studies. However, accurate detection of objects with small size, dense distribution and complex object arrangement remains a significant challenge in the remote sensing field. To address this, a refined and efficient object-detection algorithm (RE-YOLO) has been proposed in this paper for remote sensing images. Initially, a refined and efficient module (REM) was designed to balance computational complexity and feature-extraction capabilities, which serves as a key component of the RE_CSP block. RE_CSP block efficiently extracts multi-scale information, overcoming challenges posed by complex backgrounds. Moreover, the spatial extracted attention module (SEAM) has been proposed in the bottleneck of backbone to promote representative feature learning and enhance the semantic information capture. In addition, a three-branch path aggregation network (TBPAN) has been constructed as the neck network, which facilitates comprehensive fusion of shallow positional information and deep semantic information across different channels, enabling the network with a robust ability to capture contextual information. Extensive experiments conducted on two large-scale remote sensing datasets, DOTA-v1.0 and SCERL, demonstrate that the proposed RE-YOLO outperforms state-of-the-art other object-detection approaches and exhibits a significant improvement in generalization ability.

The precise thermal control of aerial cameras is crucial for the acquisition of high-resolution imagery, and an accurate temperature prediction is essential to achieve this. This paper presents a methodology for modifying thermal network models to improve the accuracy of temperature prediction for aerial cameras. Seven types of thermal parameters are extracted from the thermal network model, and a thermally sensitive analysis identifies eleven key parameters to streamline the processing time. Departing from traditional methods that rely on steady-state data, this study conducts transient thermal tests and leverages polynomial fitting to facilitate thorough parameter modification. To ensure data reliability, the Monte-Carlo algorithm is employed to explore the parameter spaces of key parameters, analyzing temperature errors. Subsequently, the Least-Squares method is utilized to obtain optimal estimates of the key parameter values. As a result, the updated model demonstrates significantly improved accuracy in temperature predictions, achieving a reduction in the maximum absolute error between the predicted and experimental results from 22 °C to 4 °C, and a lowering of the relative error from 33.8% to 6.1%. The proposed modification method validates its effectiveness in modeling and enhancing the precision of thermal network models for aerial cameras.

The quality of aerial remote sensing imaging is heavily impacted by the thermal distortions in optical cameras caused by temperature fluctuations. This paper introduces a lumped parameter thermal network (LPTN) model for the optical system of aerial cameras, aiming to serve as a guideline for their thermal design. By optimizing the thermal resistances associated with convection and radiation while considering the camera’s unique internal architecture, this model endeavors to improve the accuracy of temperature predictions. Additionally, the proposed LPTN framework enables the establishment of a heat leakage network, which offers a detailed examination of heat leakage paths and rates. This analysis offers valuable insights into the thermal performance of the camera, thereby guiding the refinement of heating zones and the development of effective active control strategies. Operating at a total power consumption of 26 W, the thermal system adheres to the low-power limit. Experimental data from thermal tests indicate that the temperatures within the optical system are maintained consistently between 19 °C and 22 °C throughout the flight, with temperature gradients remaining below 3 °C, satisfying the temperature requirements. The proposed LPTN model exhibits swiftness and efficacy in determining thermal characteristics, significantly facilitating the thermal design process and ensuring optimal power allocation for aerial cameras.

This paper addresses the problem of the two-dimensional direction-of-arrival (2D DOA) estimation of low-elevation or non-low-elevation targets using L-shaped uniform and sparse arrays by analyzing the signal models’ features and their mapping to 2D DOA. This paper proposes a 2D DOA estimation algorithm based on the dilated convolutional network model, which consists of two components: a dilated convolutional autoencoder and a dilated convolutional neural network. If there are targets at low elevation, the dilated convolutional autoencoder suppresses the multipath signal and outputs a new signal covariance matrix as the input of the dilated convolutional neural network to directly perform 2D DOA estimation in the absence of a low-elevation target. The algorithm employs 3D convolution to fully retain and extract features. The simulation experiments and the analysis of their results revealed that for both L-shaped uniform and L-shaped sparse arrays, the dilated convolutional autoencoder could effectively suppress the multipath signals without affecting the direct wave and non-low-elevation targets, whereas the dilated convolutional neural network could effectively achieve 2D DOA estimation with a matching rate and an effective ratio of pitch and azimuth angles close to 100% without the need for additional parameter matching. Under the condition of a low signal-to-noise ratio, the estimation accuracy of the proposed algorithm was significantly higher than that of the traditional DOA estimation.

The navigation accuracy of the inertial navigation system (INS) can be greatly improved when the inertial measurement unit (IMU) is effectively calibrated and compensated, such as gyro drifts and accelerometer biases. To reduce the requirement for turntable precision in the classical calibration method, a continuous dynamic self-calibration method based on a three-axis rotating frame for the hybrid inertial navigation system is presented. First, by selecting a suitable IMU frame, the error models of accelerometers and gyros are established. Then, by taking the navigation errors during rolling as the observations, the overall twenty-one error parameters of hybrid inertial navigation system (HINS) are identified based on the calculation of the intermediate parameter. The actual experiment verifies that the method can identify all error parameters of HINS and this method has equivalent accuracy to the classical calibration on a high-precision turntable. In addition, this method is rapid, simple and feasible.