On-Device Scalable Image-Based Localization via Prioritized Cascade Search and Fast One-Many RANSAC

Tran, Ngoc-Trung; Tan, Dang-Khoa Le; Doan, Anh-Dzung; Bui, Tuan-Anh; Tan, Mengxuan; Cheung, Ngai-Man

doi:10.1109/tip.2018.2881829

Cited by 35 publications

(16 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Place recognition techniques are also related to the visual localization problem as they can be used to determine which part of a scene might be visible in a query image (Cao and Snavely, 2013; Irschara et al, 2009; Jones and Soatto, 2011; Sattler et al, 2015), thus restricting the search space for feature matching. As such, place recognition techniques are used to reduce the amount of data that has to be kept in RAM, as the regions visible in the retrieved images might be loaded from disk on demand (Arth et al, 2009; Tran et al, 2019). Yet, loading 3D points from disk results in high query latency.…”

Section: Related Workmentioning

confidence: 99%

Large-scale, real-time visual–inertial localization revisited

Lynen

Zeisl

Aiger

et al. 2020

The International Journal of Robotics Research

View full text Add to dashboard Cite

The overarching goals in image-based localization are scale, robustness, and speed. In recent years, approaches based on local features and sparse 3D point-cloud models have both dominated the benchmarks and seen successful real-world deployment. They enable applications ranging from robot navigation, autonomous driving, virtual and augmented reality to device geo-localization. Recently, end-to-end learned localization approaches have been proposed which show promising results on small-scale datasets. However, the positioning accuracy, scalability, latency, and compute and storage requirements of these approaches remain open challenges. We aim to deploy localization at a global scale where one thus relies on methods using local features and sparse 3D models. Our approach spans from offline model building to real-time client-side pose fusion. The system compresses the appearance and geometry of the scene for efficient model storage and lookup leading to scalability beyond what has been demonstrated previously. It allows for low-latency localization queries and efficient fusion to be run in real-time on mobile platforms by combining server-side localization with real-time visual–inertial-based camera pose tracking. In order to further improve efficiency, we leverage a combination of priors, nearest-neighbor search, geometric match culling, and a cascaded pose candidate refinement step. This combination outperforms previous approaches when working with large-scale models and allows deployment at unprecedented scale. We demonstrate the effectiveness of our approach on a proof-of-concept system localizing 2.5 million images against models from four cities in different regions of the world achieving query latencies in the 200 ms range.

show abstract

Section: Related Workmentioning

confidence: 99%

Large-scale, real-time visual–inertial localization revisited

Lynen

Zeisl

Aiger

et al. 2020

The International Journal of Robotics Research

View full text Add to dashboard Cite

show abstract

“…Therefore, affine transform models can satisfy geometric transform requirements in our cases [22,23]. Therefore, an affine transform based on Triangle-area representation (TAR) is utilized in fine matching [24].…”

Section: Fine Matching Strategymentioning

confidence: 99%

Satellite-Borne Optical Remote Sensing Image Registration Based on Point Features

Hou

Gao

Wang

et al. 2021

Sensors

View full text Add to dashboard Cite

Since technologies in image fusion, image splicing, and target recognition have developed rapidly, as the basis of many image applications, the performance of image registration directly affects subsequent work. In this work, for rich features of satellite-borne optical imagery such as panchromatic and multispectral images, the Harris corner algorithm is combined with the scale invariant feature transform (SIFT) operator for feature point extraction. Our rough matching strategy uses the K-D (K-Dimensional) tree combined with the BBF (Best Bin First) method, and the similarity measure is the nearest neighbor/the second-nearest neighbor ratio. Finally, a triangle-area representation (TAR) algorithm is utilized to eliminate false matches in order to ensure registration accuracy. The performance of the proposed algorithm is compared with existing popular algorithms. The experimental results indicate that for visible light and multi-spectral satellite remote sensing images of different sizes and different sources, the proposed algorithm in this work is excellent in accuracy and efficiency.

show abstract

“…With the MMLESAC plane fitting technique, we improved depth segmentation over existing MLESAC and RANSAC methods. MLESAC [31][32][33][34] follows RANSAC's [35][36][37][38][39][40] basic idea which produces hypothetical results based on consecutive marginal correspondence sets; in contrast, the other remaining correspondences are used to check the quality of the hypothesis. Although, based on the probabilistic approach, MLESAC evaluates via the random sampling hypothesis, it does not presume any such complexity in the earlier matching stage which is used to provide its data.…”

Section: Multi-objects Segmentation Using Mmlesacmentioning

confidence: 99%

Automated Sustainable Multi-Object Segmentation and Recognition via Modified Sampling Consensus and Kernel Sliding Perceptron

2020

View full text Add to dashboard Cite

Object recognition in depth images is challenging and persistent task in machine vision, robotics, and automation of sustainability. Object recognition tasks are a challenging part of various multimedia technologies for video surveillance, human–computer interaction, robotic navigation, drone targeting, tourist guidance, and medical diagnostics. However, the symmetry that exists in real-world objects plays a significant role in perception and recognition of objects in both humans and machines. With advances in depth sensor technology, numerous researchers have recently proposed RGB-D object recognition techniques. In this paper, we introduce a sustainable object recognition framework that is consistent despite any change in the environment, and can recognize and analyze RGB-D objects in complex indoor scenarios. Firstly, after acquiring a depth image, the point cloud and the depth maps are extracted to obtain the planes. Then, the plane fitting model and the proposed modified maximum likelihood estimation sampling consensus (MMLESAC) are applied as a segmentation process. Then, depth kernel descriptors (DKDES) over segmented objects are computed for single and multiple object scenarios separately. These DKDES are subsequently carried forward to isometric mapping (IsoMap) for feature space reduction. Finally, the reduced feature vector is forwarded to a kernel sliding perceptron (KSP) for the recognition of objects. Three datasets are used to evaluate four different experiments by employing a cross-validation scheme to validate the proposed model. The experimental results over RGB-D object, RGB-D scene, and NYUDv1 datasets demonstrate overall accuracies of 92.2%, 88.5%, and 90.5% respectively. These results outperform existing state-of-the-art methods and verify the suitability of the method.

show abstract

On-Device Scalable Image-Based Localization via Prioritized Cascade Search and Fast One-Many RANSAC

Cited by 35 publications

References 57 publications

Large-scale, real-time visual–inertial localization revisited

Large-scale, real-time visual–inertial localization revisited

Satellite-Borne Optical Remote Sensing Image Registration Based on Point Features

Automated Sustainable Multi-Object Segmentation and Recognition via Modified Sampling Consensus and Kernel Sliding Perceptron

Contact Info

Product

Resources

About