Mr. Aurel Constantinescu Profile

Aurel Constantinescu

at Univ Politehnica din Bucuresti

SPIE Involvement:

Author

Publications (1)

This will count as one of your downloads.

You will have access to both the presentation and article (if available).

DOWNLOAD NOW

This content is available for download via your institution's subscription. To access this item, please sign in to your personal account.

Email or Username Forgot your username?

Password Forgot your password?

Show

Keep me signed in

No SPIE account? Create an account

Proceedings Article | 2 March 2023 Paper

Comparative studies for YOLOv4 target tracking algorithm performance

Aurel Constantinescu, Rodica Claudia Constantinescu, Mihai Constantinescu

Proceedings Volume 12493, 124931J (2023) https://doi.org/10.1117/12.2642502

KEYWORDS: Target detection, Detection and tracking algorithms, Image quality, Image resolution, Sensors, Image processing, Optimal decision systems, Space operations, Network architectures, Computing systems

Read Abstract +

This paper presents a detailed experimental study on operability of a target tracking system based on YOLOv4 for target detection. This work aimed to determine multi-dimensional regions where the two parameters of YOLOv4 algorithm, detection threshold and IOU threshold, allow obtaining the maximum of performance from the detection algorithm, represented by 100% detection for the overall possible detected targets. Low values of these parameters allow more detections, but also increased the number of misclassifications, counted as wrong detections. On contrast, high values of these parameters increase the detection accuracy, but on the cost of total number of detected targets. Considering that each of these parameters can have at least 9 values, 0.1 to 0.9, resulting in 81 combinations, the identification of local and global maxima and minima and of an operational zone, in which the system performance is stable, could be considered as priorities on context of optimal use of the overall target tracking system. Given the high diversity of graphics representation for the system performance in respect with the variance of these two parameters, each of them influencing in an inter-dependent mode the overall performance of the system, the authors propose that the results of each experimental complex scenario to be consolidated in 3D graph, named decision space, one for each environmental context, in which each point is the result of system efficiency reported to a specific pair of values of the mentioned algorithm parameters. These decision spaces present in aggregated form the performance obtained in a target tracking scenario, related on a specific environment, but considered all possible parameter values. The results can be used as a method of pre calibrating the target tracking system prior of effectively using it for its main purpose.

My Library

You currently do not have any folders to save your paper to! Create a new folder below.

Folder Name

Folder Description

View contact details

UPDATE YOUR PROFILE

Is this your profile? Update it now.

Sign into your SPIE.org account

Don’t have a profile and want one?

Create an account on SPIE.org

Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks. You are receiving this notice because your organization may not have SPIE eBooks access.*

*Shibboleth/Open Athens users─please sign in to access your institution's subscriptions.

To obtain this item, you may purchase the complete book in print or electronic format on SPIE.org.

ORGANIZATIONAL
Sign in with credentials provided by your organization.

Organizational Username

Organizational Password

Show/Hide Password

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members:

Non-members: ADD TO CART

Keywords/Phrases

Search In:

Publication Years