Fast and robust homography estimation method with algebraic outlier rejection

Qi, Naixin; Sheng-xiu, Zhang; Cao, Lu; Yang, Xiaogang; Li, Chuanxiang; He, Chuangxin

doi:10.1049/iet-ipr.2017.0254

Cited by 7 publications

(7 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The two geometrical models are calculated in two parallel threads: the homography H cr is computed according to our previous work [32], and the fundamental matrix F cr is computed by the eight-point algorithm [33] inside a RANSAC scheme:…”

Section: ) Compute Two Models In Parallelmentioning

confidence: 99%

See 1 more Smart Citation

Monocular Semidirect Visual Odometry for Large-Scale Outdoor Localization

Yang

et al. 2019

IEEE Access

View full text Add to dashboard Cite

To achieve large-scale outdoor real-time localization, a semidirect visual odometry method for a monocular camera is proposed. In this method, the performance of the features from the accelerated segment test (FAST) algorithm is improved to detect corners with good tracking and distribution properties. A multiscale Lucas-Kanade (LK) approach is presented to build the correspondence between features and map points robustly and efficiently. Thereafter, a semidirect visual odometry system is put forward to achieve long-distance and large-scale localization for the ground vehicle. The main contribution of this method is that it integrates the robustness and efficiency advantages of feature-based methods and the accuracy virtue of direct methods into one visual odometry system. As a result, the proposed method can be applied to localize long-distance and large-scale outdoor scenes, without loop closing and global BA. Several experiments illustrate the performances of the improved FAST algorithm, the multiscale LK approach, and our semidirect visual odometry system. The experimental results demonstrate that our semidirect visual odometry system can be operated on the KITTI benchmark accurately and efficiently. The average computational time for a single frame is below 60 ms on a general notebook computer with a CPU.

show abstract

Section: ) Compute Two Models In Parallelmentioning

confidence: 99%

“…In this step, a new frame is captured, and a Gaussian filter is exploited to remove noise. Then, the input frame is represented by a pyramid utilizing (30), ( 31), (32), and (33). If the input frame is a key-frame, new key points are extracted by the multiscale feature detection approach.…”

Section: ) Image Preprocessingmentioning

confidence: 99%

Monocular Semidirect Visual Odometry for Large-Scale Outdoor Localization

Yang

et al. 2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…[2][3][4] A widely used approach consists of the use of point correspondences in the input and reference images. 5,6 Firstly, a specified number of corresponding points are detected in the input and reference images of a scene. Next, a homography estimation approach is performed by minimization of the reprojection error.…”

Section: Introductionmentioning

confidence: 99%

Camera pose estimation based on local image correlation

Díaz-Ramírez¹,

Gonzalez-Ruiz²,

Juarez-Salazar³

2022

Optics and Photonics for Information Processing XVI

View full text Add to dashboard Cite

Camera pose estimation is an essential task in many computer vision applications. A widely used approach for this task is given by the specification of several corresponding points in a pair of captured input and reference images. The effectiveness of these methods depends on the accuracy of the specified points and is very sensitive to outliers. This work presents an iterative method for camera pose estimation based on local image correlation. The pose of the camera is estimated by finding a homography matrix that produces the best match between local fragments of the reference image constructed around the specified points and their corresponding projective transformed fragments of the input image when using the estimated homography. The performance of the proposed method is tested by processing synthetic and experimental images.

show abstract

“…Feature-based methods have been the most used to date for homography estimation [12], [13]. Most techniques mainly focus on locating points of interest in each image using detectors based on computer vision such as the Harris…”

Section: Introductionmentioning

confidence: 99%

Aggregation of Convolutional Neural Network Estimations of Homographies by Color Transformations of the Inputs

et al. 2020

View full text Add to dashboard Cite

The standard approach to the estimation of homographies consists in the application of the RANSAC algorithm to a set of tentative matches. More recent strategies based on deep learning, namely convolutional architectures, have become available. In this work, a new algorithm for the estimation of homographies is developed. It is rooted in a convolutional neural network for homography estimation, which is provided with a range of versions of the input pair of pictures. Such versions are generated by perturbation of the color levels of the input images. Each generated pair of images yields a distinct estimation of the homography, and then the estimations are combined together to obtain a final, more robust estimation. Experiments have been designed and carried out to test the validity of our approach, including qualitative and quantitative performance measures. In particular, it is demonstrated that our approach consistently outperforms the baseline approach consisting of using the output of the homography estimation deep network for the original input pair of images. INDEX TERMS Deep convolutional neural networks, homography estimation, color transformations.

show abstract

Fast and robust homography estimation method with algebraic outlier rejection

Cited by 7 publications

References 34 publications

Monocular Semidirect Visual Odometry for Large-Scale Outdoor Localization

Monocular Semidirect Visual Odometry for Large-Scale Outdoor Localization

Camera pose estimation based on local image correlation

Aggregation of Convolutional Neural Network Estimations of Homographies by Color Transformations of the Inputs

Contact Info

Product

Resources

About