Image annotation tools and dataset: a comparative analysis in brief
Rotimi-Williams Bello, Pius Adewale Owolawi, Etienne Van Wyk, Chunling Tu · 2025
To create a large database of images with ground-truth labels that benefits research in object detection and recognition, there is a need to use the right tool. Moreover, labeled data, especially in the field of computer vision, is not easier and faster to get. To achieve a large database of image annotation, LabelMe and other comparative image annotation tools were developed. This study is a comparative analysis of image annotation tools and datasets in computer vision with more emphasis on cattle object detection and segmentation. The study reveals the competitive relationship between bounding box, mask, and polygon in advancing object detection, recognition, and classification.