Skip navigation
Please use this identifier to cite or link to this item:
Files in This Item:
File Description SizeFormat 
ARTIGO_BoundingBox-freeInstance.pdf23,76 MBAdobe PDFView/Open
Title: Bounding box-free instance segmentation using semi-supervised iterative learning for vehicle detection
Authors: Carvalho, Osmar Luiz Ferreira de
Carvalho Júnior, Osmar Abílio de
Albuquerque, Anesmar Olino de
Santana, Nickolas Castro
Guimarães, Renato Fontes
Gomes, Roberto Arnaldo Trancoso
Borges, Díbio Leandro
metadata.dc.identifier.orcid: 0000-0002-5619-8525 0000-0002-0346-1684 0000-0003-1561-7583 0000-0001-6133-6753 0000-0003-4724-4064 0000-0002-9555-043X 0000-0002-4868-0629
Assunto:: Imagens aéreas
Aprendizagem profunda
Issue Date: 21-Apr-2022
Publisher: IEEE
Citation: CARVALHO, Osmar Luiz Ferreira de et al. Bounding box-free instance segmentation using semi-supervised iterative learning for vehicle detection. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, v. 15, p. 3403 - 3420, 2022. DOI: 10.1109/JSTARS.2022.3169128. Disponível em: Acesso em: 07 jul. 2022.
Abstract: Vehicle classification is a hot computer vision topic, with studies ranging from ground-view to top-view imagery. Top-view images allow understanding city patterns, traffic management, among others. However, there are some difficulties for pixel-wise classification: most vehicle classification studies use object detection methods, and most publicly available datasets are designed for this task, creating instance segmentation datasets is laborious, and traditional instance segmentation methods underperform on this task since the objects are small. Thus, the present research objectives are as follows: first, propose a novel semisupervised iterative learning approach using the geographic information system software, second, propose a box-free instance segmentation approach, and third, provide a city-scale vehicle dataset. The iterative learning procedure considered the following: first, labeling a few vehicles from the entire scene, second, choosing training samples near those areas, third, training the deep learning model (U-net with efficient-net-B7 backbone), fourth, classifying the whole scene, fifth, converting the predictions into shapefile, sixth, correcting areas with wrong predictions, seventh, including them in the training data, eighth repeating until results are satisfactory. We considered vehicle interior and borders to separate instances using a semantic segmentation model. When removing the borders, the vehicle interior becomes isolated, allowing for unique object identification. Our procedure is very efficient and accurate for generating data iteratively, which resulted in 122 567 mapped vehicles. Metrics-wise, our method presented higher intersection over union when compared to box-based methods (82% against 72%), and per-object metrics surpassed 90% for precision and recall.
Licença:: This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see
DOI: 10.1109/JSTARS.2022.3169128
Appears in Collections:GEA - Artigos publicados em periódicos

Show full item record Recommend this item " class="statisticsLink btn btn-primary" href="/handle/10482/44145/statistics">

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.