Patch-based Stereo Direct Visual Odometry Robust to Illumination Changes

Jung, Jae Hyung; Heo, Sejong; Park, Chan Gook

doi:10.1007/s12555-018-0199-2

Cited by 7 publications

(3 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, various mobile service robots have been conducting autonomous driving to provide services to humans not only in indoor but also in outdoor environments [ 10 , 11 , 12 , 13 ]. Therefore, if the robot is not inherently safe due to small size, light weight, or low power, it needs to be able to recognize the surrounding environment, and various types of SRSs are needed to build safety functions to detect approaching objects [ 14 , 15 , 16 , 17 ].…”

Section: Srs In Outdoor Environmentmentioning

confidence: 99%

Vision-Based Safety-Related Sensors in Low Visibility by Fog

Kim

Sumi

2020

Sensors

View full text Add to dashboard Cite

Mobile service robots are expanding their use to outdoor areas affected by various weather conditions, but the outdoor environment directly affects the functional safety of robots implemented by vision-based safety-related sensors (SRSs). Therefore, this paper aims to set the fog as the environmental condition of the robot and to understand the relationship between the quantified value of the environmental conditions and the functional safety performance of the robot. To this end, the safety functions of the robot built using SRS and the requirements for the outdoor environment affecting them are described first. The method of controlling visibility for evaluating the safety function of SRS is described through the measurement and control of visibility, a quantitative means of expressing the concentration of fog, and wavelength analysis of various SRS light sources. Finally, object recognition experiments using vision-based SRS for robots are conducted at low visibility. Through this, it is verified that the proposed method is a specific and effective method for verifying the functional safety of the robot using the vision-based SRS, for low visibility environmental requirements.

show abstract

Section: Srs In Outdoor Environmentmentioning

confidence: 99%

Vision-Based Safety-Related Sensors in Low Visibility by Fog

Kim

Sumi

2020

Sensors

View full text Add to dashboard Cite

show abstract

“…Jang [9] redesigned the illumination model, and the influence of lens and image sensor on image brightness is additionally modeled, but the model assumes the illumination change of the global image, so in a more general environment, such as local image illumination change, the stability of the algorithm will decline. Jung et al [10] models the change of local illumination, divides the image into small image blocks, assumes that all pixels in the same image block have the same brightness change, and implements affine transformation on all pixels in the image block. However, if the image block size of the illumination change is smaller than the image block size set by the algorithm, the algorithm will fail.…”

Section: Related Workmentioning

confidence: 99%

An Improved Visual Odometer Based on Lucas-Kanade Optical Flow and ORB Feature

et al. 2023

View full text Add to dashboard Cite

Visual odometer is an important part of SLAM system. Visual odometer based on feature points is the current mainstream. Featured based method completes the positioning by extracting feature points and matching descriptors. However, under low texture conditions, there are few features to extract, and the cost of matching descriptors is also high. to solve this problem, we propose a visual odometer based on Lucas-Kanade(LK) optical flow method and ORB feature. The method improves the performance of the ORB algorithm and acquires more feature points in a low-texture environment. A coarse-to-fine LK method is proposed for robustly and efficiently matching feature points between two frames. Based on this, a semi-direct visual odometry is proposed to improve the performance of visual odometry in low-texture environments. The main contribution of this method is that we introduce the direct method in the feature point-based approach, which significantly improves the localization accuracy in low-texture environments. The results show that the method can accomplish localization in low-texture environments without closedloop. Experimental evaluation on two benchmark datasets shows that the method has higher accuracy and robustness in motion estimation compared to other state-of-the-art methods. We have done strict tests on EuRoC and TUM data sets, and the results show that our visual odometer has higher accuracy and robustness in low-texture environment than the state-of-art feature-based scheme.

show abstract

“…However, if the ambient light source changes, the luminosity calibration is also powerless. DVO (Direct Visual Odometry) [12] and Elastic Fusion [13] also use the combination of photometric information and depth information, but their common shortcoming is that they need GPU acceleration. Elastic-Fusion uses ICP (Iterative Closest Point) algorithm to calculate the depth error and uses the method of DSO to calculate the photometric error, and then adds the two errors as the total error.…”

Section: Introductionmentioning

confidence: 99%

Direct Visual Odometry by Fusing Luminosity and Depth Information

Yin

Zhou

Zhang

et al. 2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

In the traditional direct visual odometry, it is difficult to satisfy the photometric invariant assumption due to the influence of illumination changes in the real environment, which will lead to errors and drift. This paper proposes an improved direct visual odometry system, which combines luminosity and depth information. The algorithm proposed in this paper uses Kinect 2 to collect RGB images with the corresponding depth information, and selects points with large changes of gray gradient to construct a luminosity error function and uses the corresponding depth information to construct a depth error function. The two error functions are merged into one function and converted into the least squares function of the pose of camera, the Levenberg-Marquardt algorithm is used to solve the camera pose. Finally, the Graph optimization theory and the g2o library are used to optimize the initial pose. Experiments show that the algorithm can reduce the error to a certain extent and reduce the drift caused by illumination changes.

show abstract

Patch-based Stereo Direct Visual Odometry Robust to Illumination Changes

Cited by 7 publications

References 4 publications

Vision-Based Safety-Related Sensors in Low Visibility by Fog

Vision-Based Safety-Related Sensors in Low Visibility by Fog

An Improved Visual Odometer Based on Lucas-Kanade Optical Flow and ORB Feature

Direct Visual Odometry by Fusing Luminosity and Depth Information

Contact Info

Product

Resources

About