Radar and stereo vision fusion for multitarget tracking on the special Euclidean group

Ćesić, Josip; Marković, Ivan; Cvišić, Igor

doi:10.1016/j.robot.2016.05.001

Cited by 60 publications

(45 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In contrast to [13,14],Ćesić et al [15] proposed a decision fusion scheme on the special Euclidean group using radar and stereo vision sensor [16]. For multi-object tracking, the joint integrated probabilistic data association (JIPDA) filter [44] was used, which run on matrix Lie groups.…”

Section: Object Detection By Decision Level Fusionmentioning

confidence: 99%

See 1 more Smart Citation

Spatial Attention Fusion for Obstacle Detection Using MmWave Radar and Vision Sensor

Chang

Zhang

et al. 2020

Sensors

107

View full text Add to dashboard Cite

For autonomous driving, it is important to detect obstacles in all scales accurately for safety consideration. In this paper, we propose a new spatial attention fusion (SAF) method for obstacle detection using mmWave radar and vision sensor, where the sparsity of radar points are considered in the proposed SAF. The proposed fusion method can be embedded in the feature-extraction stage, which leverages the features of mmWave radar and vision sensor effectively. Based on the SAF, an attention weight matrix is generated to fuse the vision features, which is different from the concatenation fusion and element-wise add fusion. Moreover, the proposed SAF can be trained by an end-to-end manner incorporated with the recent deep learning object detection framework. In addition, we build a generation model, which converts radar points to radar images for neural network training. Numerical results suggest that the newly developed fusion method achieves superior performance in public benchmarking. In addition, the source code will be released in the GitHub.

show abstract

Section: Object Detection By Decision Level Fusionmentioning

confidence: 99%

“…The first fusion scheme works in decision level. The prediction results from radar and vision sensor are fused to generate the final results [12][13][14][15][16][17][18][19][20][21][22][23][24]. However, there are different kinds of detection noises involved in these two kinds of prediction results.…”

Section: Introductionmentioning

confidence: 99%

Spatial Attention Fusion for Obstacle Detection Using MmWave Radar and Vision Sensor

Chang

Zhang

et al. 2020

Sensors

107

View full text Add to dashboard Cite

show abstract

“…a) Author to whom correspondence should be addressed. Electronic mail: wangkaiwei@zju.edu.cn There are various advantages of fusing RGB-D sensor and radar sensor [14][15][16][17][18] . The fusion mutually complements the drawbacks of each sensor and maximizes the capability of object detection location and recognition in varying environments.…”

Section: Introductionmentioning

confidence: 99%

Unifying obstacle detection, recognition, and fusion based on millimeter wave radar and RGB-depth sensors for the visually impaired

Long

Wang

Cheng

et al. 2019

Review of Scientific Instruments

View full text Add to dashboard Cite

It is very difficult for visually impaired people to perceive and avoid obstacles at a distance. To address this problem, the unified framework of multiple target detection, recognition and fusion is proposed based on the sensor fusion system comprised of a low-power MMW radar and an RGB-D sensor. In this paper, Mask R-CNN and SSD network are utilized to detect and recognize the objects from color images. The obstacles depth information is obtained from the depth images using the MeanShift algorithm. The position and velocity information of the multiple target are detected by the millimeter wave radar based on the principle of frequency modulated continuous wave. The data fusion based on the Particle Filter obtains more accurate state estimation and richer information by fusing the detection results from the color images, depth images and radar data compared with using only one sensor. The experiment results show that the data fusion enriches the detection results. Meanwhile, the effective detection range is expanded compared to using only the RGB-Depth sensor. Moreover, the data fusion results keep high accuracy and stability under diverse range and illumination conditions. As a wearable system, the sensor fusion system has the characteristics of versatility, portability and cost-effectiveness.

show abstract

“…Wang et al achieve a 92% detection rate with no false alarms with a two-stage fusion of mm-wave radar and monocular camera in [131]. A stereo-camera system is combined with two radar sensors by Cesic in [132]. As explained in [133,134], a radar sensor can profit from the precise lateral accuracy of a laser scanner, while contributing with an accurate relative velocity estimate of the target vehicle and a higher robustness against climatological adversities.…”

Section: Sensor Fusion For Relative Vehicle Positioningmentioning

confidence: 99%

Survey on Ranging Sensors and Cooperative Techniques for Relative Positioning of Vehicles

Müller

2017

Sensors

202

View full text Add to dashboard Cite

Future driver assistance systems will rely on accurate, reliable and continuous knowledge on the position of other road participants, including pedestrians, bicycles and other vehicles. The usual approach to tackle this requirement is to use on-board ranging sensors inside the vehicle. Radar, laser scanners or vision-based systems are able to detect objects in their line-of-sight. In contrast to these non-cooperative ranging sensors, cooperative approaches follow a strategy in which other road participants actively support the estimation of the relative position. The limitations of on-board ranging sensors regarding their detection range and angle of view and the facility of blockage can be approached by using a cooperative approach based on vehicle-to-vehicle communication. The fusion of both, cooperative and non-cooperative strategies, seems to offer the largest benefits regarding accuracy, availability and robustness. This survey offers the reader a comprehensive review on different techniques for vehicle relative positioning. The reader will learn the important performance indicators when it comes to relative positioning of vehicles, the different technologies that are both commercially available and currently under research, their expected performance and their intrinsic limitations. Moreover, the latest research in the area of vision-based systems for vehicle detection, as well as the latest work on GNSS-based vehicle localization and vehicular communication for relative positioning of vehicles, are reviewed. The survey also includes the research work on the fusion of cooperative and non-cooperative approaches to increase the reliability and the availability.

show abstract

Radar and stereo vision fusion for multitarget tracking on the special Euclidean group

Cited by 60 publications

References 34 publications

Spatial Attention Fusion for Obstacle Detection Using MmWave Radar and Vision Sensor

Spatial Attention Fusion for Obstacle Detection Using MmWave Radar and Vision Sensor

Unifying obstacle detection, recognition, and fusion based on millimeter wave radar and RGB-depth sensors for the visually impaired

Survey on Ranging Sensors and Cooperative Techniques for Relative Positioning of Vehicles

Contact Info

Product

Resources

About