Visual-based simultaneous localization and mapping (SLAM) systems perform weakly in object tracking and map reconstruction due to the unreliable depth measurement originating from image-only data. Light Detection and Ranging (LiDAR) can be coupled to overcome the drawback of uncertain depth estimation. The prerequisite for performing data fusion is to align visual-Lidar sensors to a specific coordinate system with extrinsic pose by calibrating. The conventional extrinsic calibration frameworks either rely on markers in artificial large-size calibration boards or uncontrollable natural scenes (Fig. 2 ), limiting stability and convenience. In this paper, we have designed a novel marker pattern, A4LidarTag, composed of circular holes. The difference in depth measurement is used to encode location information. Based on A4LidarTag, the automatic extrinsic calibration framework between solid-state Lidar (SSL) and the camera is developed. The proposed framework can be implemented in close range (within 1 m) and on an A4-size calibration board. The average reprojection error resulting from Lidar point clouds projection is about 0.12pixels. Experiments show excellent efficiency and versatility in both indoor and outdoor scenes.

The random noise of complementary metal–oxide–semiconductor (CMOS) image sensor is the main factor limiting the further improvement of high-accuracy spot target detection. Accordingly, a window-adaptive centroiding method based on energy iteration is proposed in this article. The method can effectively mitigate the problem of localization performance fluctuations caused by the random noise at the low-intensity pixels within the extraction window. By analyzing the centroiding error model and the random noise of CMOS detectors, simulations are used to deduce that pixel response random noise that remains after removing systematic errors is the main factor limiting further improvement in positioning accuracy. Based on the generally applicable threshold centroiding algorithm, the influence of the pixel response within the extraction window on the centroiding accuracy is derived according to the pixel energy and the pixel location relative to the target centroid. This leads to an iterative method, which combines the pixels with better performance as a new extraction window and recalculates the target centroid. The effectiveness of the algorithm at the signal-to-noise ratio (SNR) typical for real cases is simulated and analyzed. An experimental scheme is designed for the subpixel movement of point targets with a measurement platform based on a high-precision rotary table and a star tracker to validate our algorithm. Further real star experiment is conducted to verify the effectiveness of the algorithm. The results of the experiments indicate that the proposed method can reduce the random noise effect on spot extraction accuracy.

Printed circuit board (PCB) defect detection is critical for ensuring the safety of electronic devices, especially in the space industry. Non-reference-based methods, typically the deep learning methods, suffer from a large amount of annotated data requirements and poor interpretability. In contrast, conventional reference-based methods achieve higher detection accuracy by comparing with a template image but rely on precise image alignment and face the challenge of fine defects detection. To solve the problem, we propose a novel Edge-guided Energy-based PCB Defect Detection method (EEDD). We focus on the salient edge characteristic of PCB images and regard the functional defects as contour differences and then propose a novel energy measurement method for PCB contour anomaly detection. We introduce the energy transformation using the edge information of the template and test image, then Speeded-Up Robust Features method (SURF) is used for image alignment, and finally achieve defect detection by measuring the energy anomaly score pixel by pixel with the proposed energy-based defect localization and contour flood fill methods. Our method excels in detecting multi-scale defects, particularly tiny defects, and is robust against interferences such as non-finely aligned images and edge spurs. Experiments on the DeepPCB-A dataset and our high-resolution PCB dataset (HDPCB) show that the proposed method outperforms state-of-the-art methods in PCB defect-detection tasks.

Conventional vision-based sensors face limitations such as low update rates, restricted applicability, and insufficient robustness in dynamic environments with complex object motions. Single-pixel tracking systems offer high efficiency and minimal data redundancy by directly acquiring target positions without full-image reconstruction. This paper proposes a single-pixel detection system for adaptive multi-target tracking based on the geometric moment and the exponentially weighted moving average (EWMA). The proposed system leverages geometric moments for high-speed target localization, requiring merely $\mathit{3}\mathit{N}$ measurements to resolve centroids for $\mathit{N}$ targets. Furthermore, the output values of the system are used to continuously update the weight parameters, enabling adaptation to varying motion patterns and ensuring consistent tracking stability. Experimental validation using a digital micromirror device (DMD) operating at 17.857 kHz demonstrates a theoretical tracking update rate of 1984 Hz for three objects. Quantitative evaluations under 1920 × 1080 pixel resolution reveal a normalized root mean square error (NRMSE) of 0.00785, confirming the method’s capability for robust multi-target tracking in practical applications.

Aerial vehicles commonly operate in vast environments, presenting a persistent challenge in achieving high-precision localization. The contemporary popular global positioning methods have their inherent limitations. For instance, the precision of GPS is susceptible to decline or even complete failure when the signal is disrupted or absent. Furthermore, the precision of image retrieval techniques is inadequate. The construction of 3D models is a time-consuming and storage-intensive endeavor. Additionally, scene coordinate regression necessitates retraining to adapt to varying scenarios, which presents challenges when attempting to generalize across expansive environments. Addressing these challenges, we propose a network named AirGeoNet, which integrates satellite images and semantic maps to achieve high-precision efficient localization. In the first phase, we introduce the foundation model DINOV2 to extract features from satellite and aerial images, employ VLAD for image retrieval to get coarse position, and finally, significantly enhance retrieval accuracy by combining sequential images with particle filters. Subsequently, AirGeoNet matches aerial images with semantic maps to determine the three degrees of freedom in pose, including position and orientation. The semantic maps utilized by AirGeoNet are sourced from OpenStreetMap and our self-produced QMap, and training is conducted in a supervised manner using real camera poses. Our AirGeoNet method is highly efficient, requiring only a 1546-dimensional feature vector per image for image retrieval and 240k storage for a 0.9 km² semantic map while achieving state-of-the-art accuracy with single-frame localization errors of 2.854 meters on semantically rich datasets and 11 meters in complex scenarios. Our code is publicly available at https://github.com/mxz520mxz/AirGeoNet.git.