Implicit observation model for vision aided inertial navigation of aerial vehicles using single camera vector observations

Sazdovski, Vasko; Kitanov, Andrej; Petrović, Ivan

doi:10.1016/j.ast.2014.09.019

Cited by 26 publications

(10 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Camera frame and pixel coordinate. Now, the 3D feature position, two horizontal components, and height, can be obtained using equations (15) and (16). First, the height of the featureĥ n is obtained using p n, kÀ1 in equation (15) and p n, k in equation (16) aŝ…”

Section: Measurement Modelmentioning

confidence: 99%

“…using equations (13) to (16), a transformation from the feature error to the vehicle error can be made.…”

Section: First-order Linearized Modelmentioning

confidence: 99%

“…Various researches have been studied in the visionbased TRN. [16][17][18][19][20][21][22][23][24][25][26] Merhav and Bresler 17,18 proposed a recursive image registration algorithm, which makes use of the gradient of the similarity function between sequential images. Zhang et al 19,20 utilized the DEM directly in the measurement model to avoid the 3D terrain reconstruction.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Vision-based terrain referenced navigation of aerial vehicles using an adaptive extended Kalman filter

Mok

Kim

Bang

2017

Proceedings of the Institution of Mechanical Engineers, Part G:

View full text Add to dashboard Cite

This paper addresses a vision-based terrain referenced navigation of an aircraft. A digital terrain map, in the surroundings of the aircraft, is compared with the camera measurements to estimate the aircraft position. Generally, the measurement equation in the terrain referenced navigation is highly nonlinear due to the sharp changes of terrain. Thus, the conventional extended Kalman filter could lead to unstable navigation solutions. In this paper, a new approach using an adaptive extended Kalman filter is proposed to cope up with the nonlinearity problem. A least squares method is utilized to derive the linearized measurement equations. The Jacobian matrix and sensor noise covariance are modified as a means of smoothing the sharp changes of terrain. Monte Carlo simulations verify that the proposed filter gives the stable navigation solutions, even when there is a large initial error, which is the primary reason for the filter divergence. Moreover, the proposed adaptation barely requires additional computational burden, whereas the high-order filters such as particle filter generally needs higher computational power.

show abstract

Section: Measurement Modelmentioning

confidence: 99%

“…using equations (13) to (16), a transformation from the feature error to the vehicle error can be made.…”

Section: First-order Linearized Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Vision-based terrain referenced navigation of aerial vehicles using an adaptive extended Kalman filter

Mok

Kim

Bang

2017

Proceedings of the Institution of Mechanical Engineers, Part G:

View full text Add to dashboard Cite

show abstract

“…Increasing number of practical utilizations, such as vision based robotics control [58,71,80,108,111,122], surveillance [3,21,25,38,41,89,94,125], navigation [27,32,46,53,70,90,100,121], and electronics in consumer related fields [9,40,49,65,84,113,119], provide applied scenarios for the vision based computation. Most of these applications can be technically attributed into several typical technologies with respect to vision based computation.…”

Section: Introductionmentioning

confidence: 99%

Extensions of principle component analysis with applications on vision based computing

Liu¹,

Kavakli²

2015

Multimed Tools Appl

View full text Add to dashboard Cite

This paper mainly focuses on the principle component analysis (PCA) and its applications on vision based computing. The underlying mechanism of PCA given and several significant factors, involved with subspace training are discussed theoretically in detail including principle components energy, residuals assessment, and decomposition computation. The typical extensions, including probabilistic PCA (PPCA), kernel PCA (KPCA), multi-dimensional PCA and robust PCA (RPCA), have been presented with critical analysis on both mechanisms and computations. Combining with the studies on, such as, image compression, visual tracking, image recognition and super-resolution image reconstruction, PCA and its extensions applied to computer vision are critically reviewed and evaluated on the targeted issues in each application as well as the role they played at specific tasks to the characteristics, cost and suitable situations of each PCA based vision processing technique.

show abstract

“…Deep-space autonomous navigation can achieve tasks, such as orbit determination and control, attitude orientation, and target tracking, in the event of loss of ground communication; thus, it greatly enhances the survivability of the spacecraft [1]. Autonomous navigation technology used in several representative missions and plans of deep space exploration shows that the current mainstream autonomous navigation technology of deep space exploration is based on the image information of navigation star or target celestial body obtained by optical autonomous navigation (OPNAV) sensor [2], [3].…”

Section: Introductionmentioning

confidence: 99%

Accurate and Robust Synchronous Extraction Algorithm for Star Centroid and Nearby Celestial Body Edge

et al. 2019

View full text Add to dashboard Cite

Celestial body features are important navigation information in deep space exploration. This study provides a synchronous high-precision extraction algorithm for star centroid and nearby celestial body edges for a miniaturized independent optical navigation sensor, which combines the functions of a star tracker and a navigation camera. The image is filtered by a ring filter template to eliminate the interference information of background and improve the contrast between the target and the background. The secondorder directional derivative and specific area characteristic method aim to roughly extract and distinguish the features (star centroid and the nearby celestial body edge). In local area template where feature points are located, the 1D energy deviation effect is proposed to extract the features of the two different light intensity distribution models. The accuracy and robustness of our algorithm are verified by simulation and groundbased experiments. The algorithm has certain reference significance for other types of dim target and edge detections, such as infrared detection, medical image, target measurement, and machine vision. INDEX TERMS Optical autonomous navigation, high robustness and accuracy, star centroid, nearby celestial body edge, 1D energy deviation effect. GUANGJUN ZHANG received the Ph.D. degree from the

show abstract

Implicit observation model for vision aided inertial navigation of aerial vehicles using single camera vector observations

Cited by 26 publications

References 17 publications

Vision-based terrain referenced navigation of aerial vehicles using an adaptive extended Kalman filter

Vision-based terrain referenced navigation of aerial vehicles using an adaptive extended Kalman filter

Extensions of principle component analysis with applications on vision based computing

Accurate and Robust Synchronous Extraction Algorithm for Star Centroid and Nearby Celestial Body Edge

Contact Info

Product

Resources

About