Adopted image matching techniques for aiding indoor navigation

Ramadan, Mohamed M.; Tokhey, Mohamed El; Ragab, Ayman; Fath-Allah, Tamer; Ragheb, Ahmed E.

doi:10.1016/j.asej.2021.04.029

Cited by 8 publications

(2 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The SIFT algorithm was slow, and advanced applications required a faster version. The speeded-up robust features (SURF) algorithm is dependent on the principles as the SIFT, but with some approximations to execute the method much faster [36]. Similar to the SIFT, this technique is scale-invariant and rotation-invariant.…”

Section: Methodsmentioning

confidence: 99%

Survey of indoor tracking systems using augmented reality

Shewail

Elsayed²,

Zayed³

2023

IJ-AI

View full text Add to dashboard Cite

<p><span lang="EN-US">Augmented reality overlays virtual content on the physical world, displaying location-based information more efficiently. Tracking is a trace detail of the location recorded, either by taking a reading based on a set time interval, set distance, and change in direction by more than a certain angle, or a combination of these. Tracking is divided into two types, outdoor tracking, and indoor tracking. Augmented reality added value and information to tracking applications, whether indoor or outdoor. Recently, outdoor tracking by the global positioning system (GPS) became an essential component in navigation applications. However, indoor tracking is still challenging in the augmented reality field. Most augmented reality indoor tracking use optical and sensor-based tracking, such as markers, beacons, Raspberry Pi, and route planner modules for the building. With the technology enhancement, indoor tracking started to rely on WiFi, Bluetooth, geographic information system, radiofrequency identification, and sensor chip technologies. Our paper describes the recent augmented reality tracking techniques from 2011 to 2021. The paper compares image detection algorithms with various communication technologies, disscusing the advantages and the limitations of each technology.</span></p>

show abstract

Section: Methodsmentioning

confidence: 99%

Survey of indoor tracking systems using augmented reality

Shewail

Elsayed²,

Zayed³

2023

IJ-AI

View full text Add to dashboard Cite

show abstract

“…This process enables the visual positioning of the UAV, which is crucial for achieving autonomous flight and mission execution. However, current visual positioning algorithms that rely on image matching have certain shortcomings that can limit their effectiveness in certain scenarios [14][15][16][17]. These shortcomings include: (1) Image matching algorithms often rely on comparing feature descriptors extracted from images.…”

Section: Introductionmentioning

confidence: 99%

AGCosPlace: A UAV Visual Positioning Algorithm Based on Transformer

Guo,

Zhou,

Yang

2023

Drones

View full text Add to dashboard Cite

To address the limitation and obtain the position of the drone even when the relative poses and intrinsics of the drone camera are unknown, a visual positioning algorithm based on image retrieval called AGCosPlace, which leverages the Transformer architecture to achieve improved performance, is proposed. Our approach involves subjecting the feature map of the backbone to an encoding operation that incorporates attention mechanisms, multi-layer perceptron coding, and a graph network module. This encoding operation allows for better aggregation of the context information present in the image. Subsequently, the aggregation module with dynamic adaptive pooling produces a descriptor with an appropriate dimensionality, which is then passed into the classifier to recognize the position. Considering the complexity associated with labeling visual positioning labels for UAV images, the visual positioning network is trained using the publicly available Google Street View SF-XL dataset. The performance of the trained network model on a custom UAV perspective test set is evaluated. The experimental results demonstrate that our proposed algorithm, which improves upon the ResNet backbone networks on the SF-XL test set, exhibits excellent performance on the UAV test set. The algorithm achieves notable improvements in the four evaluation metrics: R@1, R@5, R@10, and R@20. These results confirm that the trained visual positioning network can effectively be employed in UAV visual positioning tasks.

show abstract