3-D image recognition system by means of stereoscopy combined with ordinary image processing

Saneyoshi, Keiji

doi:10.1109/ivs.1994.639462

Cited by 12 publications

(5 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are two representative models: those based on 3D geometries and those based on 2D geometries. One model based on a 3D geometry is a box model [1][2] . A box model takes advantage of the fact that the general geometry of a vehicle can be approximated in the form of a rectangle and is highly versatile because it does not rely on the vehicle type, color, or other characteristics for vehicle detection.…”

Section: Vehicle Modelsmentioning

confidence: 99%

Preceding Vehicle Detection Using Stereo Images and Non-scanning Millimeter-Wave Radar

Segawa

Shiohara

Sasaki

et al. 2006

IEICE Transactions on Information and Systems

View full text Add to dashboard Cite

Eigo SEGAWA †, M or it o SHI OHARA †, Shige r u SASAKI †, Nor io HASHI GUCHI ‡, Tomonobu TAKASHI M A ‡, a nd M a s a t os hi TOHNO ‡ †Fuj it s u La bor a t or ie s Lt d., ‡Fuj it s u Limit e d, 4-1-1 Ka mikoda na ka , Na ka ha r a -ku, Ka wa s a ki, Ka na ga wa 211-8588, J a pa n. s e ga wa .e igo@ j p.f uj it s u.c om Abs t ractWe developed a system that detects the vehicle driving immediately ahead of one's own car in the same lane and measures the distance to and relativ e speed of that v ehicle to prev ent accidents such as rear-end collisions. The system is the first in the industry to use non-scanning millimeter-wav e radar combined with a sturdy stereo image sensor, which k eeps cost low. I t can operate stably in adv erse weather conditions such as rain and snow, which could not easily be done with prev ious sensors. The system' s v ehicle detection performance was tested, and the system can correctly detect v ehicles driv ing 3 to 50 m ahead in the same lane with higher than 9 9% accuracy in clear weather. Detection performance in rainy weather, where water drops and splashes notably degraded v isibility, was higher than 90%.

show abstract

Section: Vehicle Modelsmentioning

confidence: 99%

Preceding Vehicle Detection Using Stereo Images and Non-scanning Millimeter-Wave Radar

Segawa

Shiohara

Sasaki

et al. 2006

IEICE Transactions on Information and Systems

View full text Add to dashboard Cite

show abstract

“…Many researchers are focusing on developing new computer vision algorithms and techniques, such as the ones published by Subaru Research Center [3], the basis of today's EyeSight system [4]. Over the last decade, these technologies have improved considerably, due to increasing investments in autonomous vehicles, driven mostly by companies such as Google, Tesla, Volvo, and others.…”

Section: Introductionmentioning

confidence: 99%

“…These kinds of sensors are expensive, which makes their adoption in entry‐level vehicles not viable. Some researchers have also been using stereo cameras [3] and stereo correspondence algorithms [8] as depth extractor, instead of using LiDAR or radars. Then, with the rise of deep learning and convolutional neural networks (CNNs), solutions using only cameras have proven to be efficient (for object detection and depth estimation for stereo vision) and economically viable.…”

Section: Introductionmentioning

confidence: 99%

Three‐stage RGBD architecture for vehicle and pedestrian detection using convolutional neural networks and stereo vision

Ferraz

Oliveira

Ferreira

et al. 2020

IET intell. transp. syst

View full text Add to dashboard Cite

“…Most of the obstacle detection systems are based on stereo vision (Bertozzi et al, 1997(Bertozzi et al, , 2000 and use an inverse perspective model applied to a planar model (static or dynamic) of the road. Optical flow based approaches (Enkelmann, 1997) and correlation based stereo systems (Saneyoshi, 1994) have also been used. A system developed at Daimler-benz Research also uses stereo vision (Franke et al, 1999).…”

Section: Introductionmentioning

confidence: 99%

A Stereovision Method for Obstacle Detection and Tracking in Non-Flat Urban Environments

Araújo

Wang

2005

Auton Robot

View full text Add to dashboard Cite

Abstract.Obstacle detection is an essential capability for the safe guidance of autonomous vehicles, especially in urban environments. This paper presents an efficient method to integrate spatial and temporal constraints for detecting and tracking obstacles in urban environments. In order to enhance the reliability of the obstacle detection task, we do not consider the urban roads as rigid planes, but as quasi-planes, whose normal vectors have orientation constraints. Under this flexible road model, we propose a fast, robust stereovision based obstacle detection method. A watershed transformation is employed for obstacle segmentation in dense traffic conditions, even with partial occlusions, in urban environments. Finally a UKF (Unscented Kalman filter) is applied to estimate the obstacles parameters under a nonlinear observation model. To avoid the difficulty of the computation in metric space, the whole detection process is performed in the disparity image. Various experimental results are presented, showing the advantages of this method.

show abstract

3-D image recognition system by means of stereoscopy combined with ordinary image processing

Cited by 12 publications

References 1 publication

Preceding Vehicle Detection Using Stereo Images and Non-scanning Millimeter-Wave Radar

Preceding Vehicle Detection Using Stereo Images and Non-scanning Millimeter-Wave Radar

Three‐stage RGBD architecture for vehicle and pedestrian detection using convolutional neural networks and stereo vision

A Stereovision Method for Obstacle Detection and Tracking in Non-Flat Urban Environments

Contact Info

Product

Resources

About