Seed-Based Superpixel Re-Segmentation for Improving Object Delineation

Autor(es)

Lacerda, Lucca S. P. ; Belem, Felipe C. ; Goncalves do Patrocinio Junior, Zenilton Kleber ; Falcao, Alexandre X. ; Guimaraes, Silvio J. F.

• Tipo de documento: Artigo Científico
• Fonte: PROGRESS IN PATTERN RECOGNITION, IMAGE ANALYSIS, COMPUTER VISION, AND APPLICATIONS, CIARP 2024, PT I; v. 15368, p. 14-pg., 2025-01-01.

Resumo

Superpixels are an effective image segmentation strategy, whose results apply and assist in classification tasks. However, by aiming for maximum performance with a minimum quantity of regions, the object delineation may be compromised, demanding re-segmentation. In this paper, we propose and evaluate three re-segmentation strategies that rely on a novel and accurate superpixel framework, named SICLE, and require minimal intervention from the user (i.e., up to three clicks). Our qualitative and quantitative results show significant improvement over the previous superpixel segmentation for separating the object of interest from the background. (AU)

• Processo FAPESP: 23/14427-8 - Ciência de Dados para a Indústria Inteligente (CDII)
• Beneficiário: José Alberto Cuminato
• Modalidade de apoio: Auxílio à Pesquisa - Programa Centros de Pesquisa Aplicada