Sensor Fusion of Laser &amp; Stereo Vision Camera for Depth Estimation and Obstacle Avoidance

Kumar, Suresh; Gupta, Daya; Yadav, Sakshi

doi:10.5120/485-795

Cited by 19 publications

(10 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(2) (3) and : Upper and lower bound of RCS for pedestrian and : Upper and lower bound of RCS for vehicle Furthermore, if the object distance or area irradiated by the radio waves varies, G t , G r , and R will also change. It is thought, however, that many other change factors originated from a change in the pedestrian and vehicle's RCS noted as ρ p and ρ v , respectively.…”

Section: Radar Cross Section (Rcs)mentioning

confidence: 99%

“…Two of such technologies are the city safety system and collision warning with full auto brake system developed by Volvo. In early stage, these systems used single sensor such as the camera or radar to find objects for further alert [3,4,5]. However, these systems cannot provide accurate surrounding localization due to their inherent limitations.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Pedestrian and Vehicle Recognition Based on Radar for Autonomous Emergency Braking

Wu¹

2017

SAE Technical Paper Series

View full text Add to dashboard Cite

Autonomous Emergency Braking Systems (AEBS) usually contain radar, (stereo) camera and/or LiDAR-based technology to identify potential collision partners ahead of the car, such that to warn the driver or automatically brake to avoid or mitigate a crash. The advantage of camera is less cost: however, is inevitable to face the defects of cameras in AEBS, that is, the image recognition cannot perform good accuracy in the poor or over-exposure light condition. Therefore, the compensation of other sensors is of importance. Motivated by the improvement of false detection, we propose a Pedestrian-and-Vehicle Recognition (PVR) algorithm based on radar to apply to AEBS. The PVR employs the radar cross section (RCS) and standard deviation of width of obstacle to determine whether a threshold value of RCS and standard deviation of width of the pedestrian and vehicle is crossed, and to identity that the objective is a pedestrian or vehicle, respectively. The performance of the proposed algorithm is pressed via the experimental test data.

show abstract

Section: Radar Cross Section (Rcs)mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Pedestrian and Vehicle Recognition Based on Radar for Autonomous Emergency Braking

Wu¹

2017

SAE Technical Paper Series

View full text Add to dashboard Cite

show abstract

“…Sensor fusion systems can make use of cameras with depth information like Kinect [16] and RealSense [17]. Additionally, some related works make use of stereo cameras to track a target, in which both cameras are placed at a predetermined distance and rotation, calculating a whole 3D reconstruction of the scene [18,19,20] using Epipolar Geometry [21]. A single-camera system can also be deployed [22] simulating a stereo system, either using markers [23], or previously establishing the separation parameters among two images [24], knowing the relationship between the key-points of both images, necessary to build the Epipolar Geometry.…”

Section: Introductionmentioning

confidence: 99%

Monocular Robust Depth Estimation Vision System for Robotic Tasks Interventions in Metallic Targets

Almagro

Castro

Lunghi

et al. 2019

Sensors

View full text Add to dashboard Cite

Robotic interventions in hazardous scenarios need to pay special attention to safety, as in most cases it is necessary to have an expert operator in the loop. Moreover, the use of a multi-modal Human-Robot Interface allows the user to interact with the robot using manual control in critical steps, as well as semi-autonomous behaviours in more secure scenarios, by using, for example, object tracking and recognition techniques. This paper describes a novel vision system to track and estimate the depth of metallic targets for robotic interventions. The system has been designed for on-hand monocular cameras, focusing on solving lack of visibility and partial occlusions. This solution has been validated during real interventions at the Centre for Nuclear Research (CERN) accelerator facilities, achieving 95% success in autonomous mode and 100% in a supervised manner. The system increases the safety and efficiency of the robotic operations, reducing the cognitive fatigue of the operator during non-critical mission phases. The integration of such an assistance system is especially important when facing complex (or repetitive) tasks, in order to reduce the work load and accumulated stress of the operator, enhancing the performance and safety of the mission.

show abstract

“…There are three approaches for image-based distance computation namely, stereovision, monovision, and time-of-flight. Stereo vision uses two similar active cameras [9] separated by a distance in either a standalone fashion or eye-in-hand configuration for 3D visual servoing to find the disparity map and depth. This method [7,10] gives high accuracy, but implementing this technique is expensive (as it requires two cameras) and requires high computation time due to simultaneous processing of many images.…”

Section: Introductionmentioning

confidence: 99%

Shadow Removal in Acquired Image for Visual Servoing of Unmanned Ground Vehicles

K¹,

Parthasarathy²,

Prathibha³

2018

IJIES

View full text Add to dashboard Cite

Vision-guided robotic operation is one of the new concepts in teleoperated Unmanned Ground Vehicle (UGV) for military applications. The objective of the Visual Servoing (VS) is to control the position of the robotic arm using the data such as the distance of an object from the reference frame, or the length and width of the object extracted from the vision sensor. In Image-Based Visual Servoing (IBVS) scheme, position control values are computed from image features directly. This work proposes the technique of shadow removal of the object through the advanced light model from the image, acquired using a monocular camera. The 2D spatial coordinates and Pointof-Contact (PoC) of the object with reference to the ground plane is computed using straight-line equations. PoC analysis is made through mapping of pixel distances to spatial distances to analyses mean pixel distance, and the percentage of error obtained between actual distance and calculated distance computed. Other parameters like the width of the object and, the distance between the camera and object is also estimated. The computed result yields a maximum error of 3 cm and shows 2.6% error for camera fixed at the height of 160 cm with a tilt angle of 50˚, which covers an area of 150 cm x 120 cm.

show abstract

Sensor Fusion of Laser & Stereo Vision Camera for Depth Estimation and Obstacle Avoidance

Cited by 19 publications

References 14 publications

Pedestrian and Vehicle Recognition Based on Radar for Autonomous Emergency Braking

Pedestrian and Vehicle Recognition Based on Radar for Autonomous Emergency Braking

Monocular Robust Depth Estimation Vision System for Robotic Tasks Interventions in Metallic Targets

Shadow Removal in Acquired Image for Visual Servoing of Unmanned Ground Vehicles

Contact Info

Product

Resources

About