Efficient Feature Distribution for Object Matching in Visual-Sensor Networks

Sulič, Vildana; Kristan, Matej; Kovačič, Stanislav

doi:10.1109/tcsvt.2011.2133330

Cited by 12 publications

(6 citation statements)

References 35 publications

(34 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…S MART camera networks are composed of nodes that perform image processing locally and aim to transfer the minimum amount of data over the network to accomplish collaborative tasks such as object detection and tracking [1], [2]. The challenges involved in smart camera networks include reducing the amount of data to be processed and shared across the network, real-time operation and energy efficiency [3], [4].…”

Section: Introductionmentioning

confidence: 99%

Person Re-identification

Bąk¹,

Brémond²

2021

Computer Vision

View full text Add to dashboard Cite

Networks of smart cameras share large amounts of data to accomplish tasks such as re-identification. We propose a feature selection method that minimizes the data needed to represent the appearance of objects by learning the most appropriate feature set for the task at hand (person re-identification). The computational cost for feature extraction and the cost for storing the feature descriptor are considered jointly with feature performance in order to select cost-effective good features. This selection allows us to improve inter-camera re-identification while reducing the bandwidth that is necessary to share data across the camera network. We also rank the selected features in the order of effectiveness for the task to enable a further reduction of the feature set by dropping the least effective features when application constraints require this adaptation. We compare the proposed approach with state-of-the-art methods on the i-LIDS and VIPeR datasets and show that the proposed approach considerably reduces network traffic due to inter-camera feature sharing while keeping the re-identification performance at an equivalent or better level compared with the state of the art.

show abstract

Section: Introductionmentioning

confidence: 99%

Person Re-identification

Bąk¹,

Brémond²

2021

Computer Vision

View full text Add to dashboard Cite

show abstract

“…Object matching is a fundamental task in multi-view and multicamera scene observation [1,2,3,4,5]. Existing approaches generally focus on object matching accuracy without considering constraints on the available resources, such as storage, battery life, communication and computational capabilities [6], which are important for battery-powered devices such as smartphones and wireless smart cameras.…”

Section: Introductionmentioning

confidence: 99%

Low-cost multi-camera object matching

Tahir

Cavallaro

2014

2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

We propose an object matching approach aimed at smartphone cameras that exploits the well-known concept of local sets of features for object representation. We also enable the temporal alignment of cameras by exploiting the frames of detected objects to group objects appeared in the same time interval for the assignment within each camera. The proposed approach does not need training thus making it suitable for matching during short temporal intervals. We use both outdoor and indoor datasets for the evaluation, and show that the proposed method reduces up to 95% the amount of information to be stored and communicated.

show abstract

“…In object detection and recognition, feature matching plays an important role in identifying similar objects seen from different views or angles. Some of the applications that use feature matching techniques are image stitching [1], fingerprint verification [2], tracking [3] and visual surveillance [4]. In order to reliably identify an object, the selected features must be robust to scale variance, noise, rotation, viewpoint changes and illumination changes.…”

Section: Introductionmentioning

confidence: 99%

Enhanced Rotational Feature Points Matching using Orientation Correction

Goh

Abu-Bakar

Mokji

et al. 2014

ELCVIA

View full text Add to dashboard Cite

In matching between images, several techniques have been developed particularly for estimating orientation assignment in order to make feature points invariant to rotation. However, imperfect estimation of the orientation assignment may lead to feature mismatching and a low number of correctly matched points. Additionally, several possible candidates with high correlation values for one feature in the reference image may lead to matching confusion. In this paper, we propose a post-processing matching technique that will not only increase the number of correctly matched points but also manage to solve the above mentioned two issues. The key idea is to modify feature orientation based on the relative rotational degree between two images, obtained by taking the difference between the major correctly matched points in the first matching cycle. From the analysis, our proposed method shows that the number of detected points correctly matched with the reference image can be increased by up to 50%. In addition, some mismatched points due to similar correlation values in the first matching round can be corrected. Another advantage of the proposed algorithm it that it can be applied to other state-of-the-art orientation assignment techniques.

show abstract

Efficient Feature Distribution for Object Matching in Visual-Sensor Networks

Cited by 12 publications

References 35 publications

Person Re-identification

Person Re-identification

Low-cost multi-camera object matching

Enhanced Rotational Feature Points Matching using Orientation Correction

Contact Info

Product

Resources

About