Density Enhancement-Based Long-Range Pedestrian Detection Using 3-D Range Data

Li, Keqiang; Wang, Xiao; Xu, Youchun; Wang, Jianqiang

doi:10.1109/tits.2015.2502325

Cited by 52 publications

(24 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Three-dimensional features from point clouds and two-dimensional features from projection images were extracted to classify pedestrians with a radial basis function kernel SVM [13]. In order to detect pedestrians in a long-range area with sparse point clouds, radial basis function-based interpolation was implemented to robustly extract features and SVMs are applied to learn the classifiers [14].…”

Section: Training-based Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Pedestrian Detection with Lidar Point Clouds Based on Single Template Matching

Liu

Wang

2019

Electronics

View full text Add to dashboard Cite

In the field of intelligent transportation systems, pedestrian detection has become a problem that is urgently in need of a solution. Effective pedestrian detection reduces accidents and protects pedestrians from injuries. A pedestrian-detection algorithm, namely, single template matching with kernel density estimation clustering (STM-KDE), is proposed in this paper. First, the KDE-based clustering method is utilized to extract candidate pedestrians in point clouds. Next, the coordinates of the point clouds are transformed into the pedestrians’ local coordinate system and projection images are generated. Locally adaptive regression kernel features are extracted from the projection image and matched with the template features by using cosine similarity, based on which pedestrians are distinguished from other columnar objects. Finally, comparative experiments using KITTI datasets are conducted to verify pedestrian-detection performance. Compared with the STM with radially bounded nearest neighbor (STM-RBNN) algorithm and the KDE-based pedestrian-detection algorithm, the proposed algorithm can segment gathering pedestrians and distinguish them from other columnar objects in real scenarios.

show abstract

Section: Training-based Methodsmentioning

confidence: 99%

“…(x, y, z), (x , y , z ) and (x , y , z ) represent the coordinates in Oxyz, O x y z and Ox y z , respectively. The coordinate transformation from Lidar coordinate system Oxyz to local coordinate system O x y z is as follows [14].…”

Section: Multilayer Fusionmentioning

confidence: 99%

Pedestrian Detection with Lidar Point Clouds Based on Single Template Matching

Liu

Wang

2019

Electronics

View full text Add to dashboard Cite

show abstract

“…Different from rigid objects, non-rigid objects are deformable and may change shape during tracking. Therefore, for non-rigid object tracking, especially for person tracking, early works usually depend on trained classifier [9,27,28] to detect objects even if their shape have changed. After getting detection results from the detector every frame, they usually use filter methods [5,6] to match the detection results, which can track objects.…”

Section: Object Tracking Using 3d Point Cloudsmentioning

confidence: 99%

Point Siamese Network for Person Tracking Using 3D Point Clouds

Cui

Fang

Zhou

2019

Sensors

View full text Add to dashboard Cite

Person tracking is an important issue in both computer vision and robotics. However, most existing person tracking methods using 3D point cloud are based on the Bayesian Filtering framework which are not robust in challenging scenes. In contrast with the filtering methods, in this paper, we propose a neural network to cope with person tracking using only 3D point cloud, named Point Siamese Network (PSN). PSN consists of two input branches named template and search, respectively. After finding the target person (by reading the label or using a detector), we get the inputs of the two branches and create feature spaces for them using feature extraction network. Meanwhile, a similarity map based on the feature space is proposed between them. We can obtain the target person from the map. Furthermore, we add an attention module to the template branch to guide feature extraction. To evaluate the performance of the proposed method, we compare it with the Unscented Kalman Filter (UKF) on 3 custom labeled challenging scenes and the KITTI dataset. The experimental results show that the proposed method performs better than UKF in robustness and accuracy and has a real-time speed. In addition, we publicly release our collected dataset and the labeled sequences to the research community.

show abstract

“…3D LIDAR is one of the most prevalent sensors used in the autonomous vehicle perceptual systems, and it has a wide range of view, with precise depth information, and long-range and night-vision capabilities in target recognition [2][3][4]. In the object detection task, 3D LIDAR has certain advantages over cameras in acquiring the pose and shape of the detected objects, since laser scans contain spatial coordinates of the point clouds by nature [5].…”

Section: Introductionmentioning

confidence: 99%

Fusion of 3D LIDAR and Camera Data for Object Detection in Autonomous Vehicle Applications

et al. 2020

View full text Add to dashboard Cite

It's critical for an autonomous vehicle to acquire accurate and real-time information of the objects in its vicinity, which will fully guarantee the safety of the passengers and vehicle in various environment. 3D LIDAR can directly obtain the position and geometrical structure of the object within its detection range, while vision camera is very suitable for object recognition. Accordingly, this paper presents a novel object detection and identification method fusing the complementary information of two kind of sensors. We first utilize the 3D LIDAR data to generate accurate object-region proposals effectively. Then, these candidates are mapped into the image space where the regions of interest (ROI) of the proposals are selected and input to a convolutional neural network (CNN) for further object recognition. In order to identify all sizes of objects precisely, we combine the features of the last three layers of the CNN to extract multi-scale features of the ROIs. The evaluation results on the KITTI dataset demonstrate that : (1) Unlike sliding windows that produce thousands of candidate object-region proposals, 3D LIDAR provides an average of 86 real candidates per frame and the minimal recall rate is higher than 95%, which greatly lowers the proposals extraction time; (2) The average processing time for each frame of the proposed method is only 66.79ms, which meets the real-time demand of autonomous vehicles; (3) The average identification accuracies of our method for car and pedestrian on the moderate level are 89.04% and 78.18% respectively, which outperform most previous methods.

show abstract

Density Enhancement-Based Long-Range Pedestrian Detection Using 3-D Range Data

Cited by 52 publications

References 34 publications

Pedestrian Detection with Lidar Point Clouds Based on Single Template Matching

Pedestrian Detection with Lidar Point Clouds Based on Single Template Matching

Point Siamese Network for Person Tracking Using 3D Point Clouds

Fusion of 3D LIDAR and Camera Data for Object Detection in Autonomous Vehicle Applications

Contact Info

Product

Resources

About