Robust Matching Cost Function for Stereo Correspondence Using Matching by Tone Mapping and Adaptive Orthogonal Integral Image

Dinh, Vinh Quang; Nguyen, Vinh Dinh; Jeon, Jae Wook

doi:10.1109/tip.2015.2481702

Cited by 11 publications

(10 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The first step is to calculate the matching costs using Zero Mean Normalized Cross-Correlation (ZNCC) [60], which compares the structure of two zero-mean signals. In contrast to other metrics, e.g., SAD, SSD, or Census [58], the ZNCC is substantially more robust for the CAMSI setup as the image content differs significantly for multi-spectral images. For two vectors a and b, ZNCC is defined as To assess the quality of the estimation, a search is also performed in the opposite direction from the peripheral view k to the center view, which leads to the cost matrix…”

Section: B Fast One-dimensional Disparity Estimationmentioning

confidence: 90%

“…II) to perform pairwise registrations of the center view and the various peripheral views. Altogether, neither widely-distributed template matching-approaches, e.g., [58], nor deep learning methods [59] provide a satisfactory solution due to the large deviation of the image content. For deep learning algorithms, another problem is the limited amount of multi-spectral, multi-view training data, which restricts a reasonable application.…”

Section: B Fast One-dimensional Disparity Estimationmentioning

confidence: 99%

“…The highlighted regions are used to calculate an average depth µ z as well as depth deviation σ z using the relationship shown in (3). This procedure is repeated by generating further disparity maps by applying two state-of-the-art cross-spectral stereo matching algorithms, namely Census + SGM [58], [64], [44] and CCNG+SGM [47], [64]. The resulting disparity maps are averaged, so that a combined map is generated, which takes all camera viewpoints into account.…”

Section: A Registration Performancementioning

confidence: 99%

See 2 more Smart Citations

Camera Array for Multi-Spectral Imaging

Genser

Seiler

Kaup

2020

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

Recently, many new applications arose for multispectral and hyper-spectral imaging. Besides modern biometric systems for identity verification, also agricultural and medical applications came up, which measure the health condition of plants and humans. Despite the growing demand, the acquisition of multi-spectral data is up to the present complicated. Often, expensive, inflexible, or low resolution acquisition setups are only obtainable for specific professional applications. To overcome these limitations, a novel camera array for multi-spectral imaging is presented in this article for generating consistent multispectral videos. As differing spectral images are acquired at various viewpoints, a geometrically constrained multi-camera sensor layout is introduced, which enables the formulation of novel registration and reconstruction algorithms to globally set up robust models. On average, the novel acquisition approach achieves a gain of 2.5 dB PSNR compared to recently published multi-spectral filter array imaging systems. At the same time, the proposed acquisition system ensures not only a superior spatial, but also a high spectral, and temporal resolution, while filters are flexibly exchangeable by the user depending on the application. Moreover, depth information is generated, so that 3D imaging applications, e.g., for augmented or virtual reality, become possible. The proposed camera array for multi-spectral imaging can be set up using off-the-shelf hardware, which allows for a compact design and employment in, e.g., mobile devices or drones, while being cost-effective.

show abstract

Section: B Fast One-dimensional Disparity Estimationmentioning

confidence: 90%

Section: B Fast One-dimensional Disparity Estimationmentioning

confidence: 99%

Section: A Registration Performancementioning

confidence: 99%

See 1 more Smart Citation

Camera Array for Multi-Spectral Imaging

Genser

Seiler

Kaup

2020

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

show abstract

“…Recently, the linkage artificial bee colony (LABC) method [5] has been introduced to address the linkage problem where variables are dependent and interactive. However, LABC does not employ important available information from the problem models, such as the Poss models [6] in image processing and computer vision that a pixel (a variable) interacts locally with its neighboring pixels.…”

Section: Introductionmentioning

confidence: 99%

Neighbor-Interactive Bee Colony for Problems with Local Structures

Hong

Ahn²,

Jeong

2017

IEICE Trans. Fundamentals

View full text Add to dashboard Cite

In this letter, we integrate domain information into the original artificial bee colony algorithm to create a novel, neighborinteractive bee colony algorithm. We use the Hamming distance measure to compute variable dependency between two binary variables and employ the Gini correlation coefficient to compute variable relation between integer variables. The proposed optimization method was evaluated by minimizing binary Ising models, integer Potts models, and trapped functions. Experimental results show that the proposed method outperformed the traditional artificial bee colony and other meta-heuristics in all the testing cases.

show abstract

“…However, the majority of existing stereo matching functions exploit the color consistent assumption and use simple matching cost functions [5]- [7], so their performances are degraded severely in real stereo images whose intensities are arbitrarily changed [8].…”

Section: Introductionmentioning

confidence: 99%

Statistics-Based Matching Cost Function for Intensity Transformation

Hong¹,

Ahn²

2016

IJCCE

View full text Add to dashboard Cite

Stereo matching is a challenging task due to stereo images being affected by many factors such as radiometric distortion, sun and rain flare, flying snow, occlusion and object boundaries. However, most of the existing stereo matching methods assume that corresponding pixels in left and right images have the same intensity; accordingly, the stereo matching methods use simple matching cost functions. As a result, they degrade their performance significantly when operating with real-world stereo images whose intensities of corresponding pixels can be arbitrarily transformed. State-of-the-art matching cost functions under conditions of distorted intensities between stereo images, such as the census transform, adaptive normalized cross-correlation, and the support local binary pattern perform with limited accuracy. In this paper, we propose a novel matching cost function based on a robust order statistic coefficient and segmentation that can operate accurately under various conditions of transformed intensities between stereo images. Based on the robust order statistic coefficient, the proposed matching cost function can tolerate local, monotonically nonlinear changes in intensities between the left and right images. By using the information of segmentation, the matching cost function operates accurately at object boundaries. The qualitative and quantitative experimental results obtained using stereo images in different datasets under various conditions show that our proposed matching cost function outperforms state-of-the-art matching cost functions in indoor and outdoor stereo images with various radiometric distortions.

show abstract

Robust Matching Cost Function for Stereo Correspondence Using Matching by Tone Mapping and Adaptive Orthogonal Integral Image

Cited by 11 publications

References 27 publications

Camera Array for Multi-Spectral Imaging

Camera Array for Multi-Spectral Imaging

Neighbor-Interactive Bee Colony for Problems with Local Structures

Statistics-Based Matching Cost Function for Intensity Transformation

Contact Info

Product

Resources

About