Year: 2026
Type: article
Abstract: The headway of convolutional neural systems (CNNs) and Deep learning (DL) in the previous decade brought about significant upgrades in computer vision. One of the recipients of these advances is the task of object detection, where the goal is to distinguish and locate real-world objects inside pictures or videos. Real-time object tracking on a drone under a dynamic situation has been a difficult issue for a long time, with existing methodologies utilizing off-line calculation or powerful computation units on board. This paper displays lightweight real-time on board object tracking methodology, which varies, from basic image classification in that the AI demonstrate needs to distinguish numerous objects in a single frame, and figure out where these objects are found. The advances in procedures, joined with the improved PC equipment, put real-time object detection well inside the capacities of present day processors. Real-time object recognition is essential for some utilization of Unmanned Aerial Vehicles (UAVs), for example, observation and reconnaissance, search-and-rescue, and foundation assessment. In the previous couple of years, Convolutional Neural Networks (CNNs) have ascended as an unbelievable class of models for recognizing picture content, and are seen as the standard strategy for generally issue
Institution: REVA University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7216862929
