Long-distance multi-vehicle detection at night is critical in military operations. Due to insufficient light at night, the visual features of vehicles are difficult to distinguish, and many missed detections occur. This paper proposes a two-level detection method for long-distance nighttime multi-vehicles based on Gm-APD lidar intensity images and point cloud data. The method is divided into two levels. The first level is 2D detection, which enhances the local contrast of the intensity image and improves the brightness of weak and small objects. With the confidence threshold set, the detection result greater than the threshold is reserved as a reliable object, and the detection result less than the threshold is a suspicious object. In the second level of 3D recognition, the suspicious object area from the first level is converted into the corresponding point cloud classification judgment, and the object detection score is obtained through comprehensive judgment. Finally, the object results of the two-level recognition are merged into the final detection result. Experimental results show that the method achieves a detection accuracy of 96.38% and can effectively improve the detection accuracy of multiple vehicles at night, which is better than the current state-of-the-art detection methods.

No abstract

Geiger mode Avalanche Photo Diode (Gm-APD) array lidar is a lidar that can perform single-photon detection. It offers a wide range of applications due to its low power consumption, small size, and extended detecting distance. There haven't been many research on this detector's target classification because of its late development and small detector array. The classification technique based on the Gm-APD array lidar point cloud is the focus of this paper's research: Firstly, the Gm-APD array lidar is utilized to perform imaging tests on four targets from various angles in order to create a target classification dataset.Following that, several data preprocessing methods were chosen and implemented based on the characteristics of the obtained data, such as filling in missing values, performing range image and intensity image interpolation, using the principle of keyhole imaging to convert the range image to point cloud data, realizing the information fusion of distance image and intensity image, and using multiple point cloud data enhancement methods. Finally, the point cloud classification networks PointNet and PointNet++ are trained on point cloud data with varying levels of preprocessing, the results are compared and analyzed, and the impact of different preprocessing methods on the classification accuracy of the two networks is determined. Inferences were made and experiments were carried out to verify the inferences. The data set preprocessing method with the highest classification accuracy of the two networks is discovered, laying the groundwork for future Gm-APD lidar target classification and detection research.