IMAGE SEGMENTATION FOR PEOPLE IDENTIFICATION: ANEVALUATION OF UNSUPERVISED TECHNIQUES

SEGMENTAÇÃO DE IMAGENS PARA IDENTIFICAÇÃO DE PESSOAS: UMA AVALIAÇÃO DE TÉCNICAS NÃO SUPERVISIONADAS

Santos, Lucas Lisboa dos; Pagano, Tiago; Vacaro, Juliano; Loureiro, Rafael; Junior, Neilton; Cunha, Guilherme da; Nascimento, Erick Giovani Sperandio; Winkler, Ingrid

Artigo completo:

The evaluation of segmentation techniques is a complex activity since itdepends on the target purpose. Our research is a technical evaluation ofsegmentation, specifically, it aims to evaluate the techniques Ant Colony FuzzyC-means Hybrid Algorithm (AFHA), Region Splitting and Merging Fuzzy C-meansHybrid Algorithm (RFHA) with the distance between points and Kanezaki, to identifypeople in images from the perspective of Jaccard Index and F Measure metrics(J&F). The method was divided into four stages: the selection of the image sample,evaluation process, experiment execution, and results composed by segmentedimage, group, and J&F metrics. The results indicate Kanezaki has surpassed theother techniques. It is recommended future research to identify whether a correlationbetween quantitative and qualitative analysis exists.

A avaliação das técnicas de segmentação é uma atividade complexa, pois depende do objetivo da segmentação. Nossa pesquisa é uma avaliação de técnicas de segmentação, mais especificamente ela tem como objetivo avaliar as técnicas Ant Colony Fuzzy C-means Hybrid Algorithm (AFHA), Region Splitting and Merging Fuzzy C-means Hybrid Algorithm (RFHA) com variações na distância entre pontos e Kanezaki, para identificar pessoas em imagens sob perspectiva das métrica métricas Jaccard Index e F Measure (J&F). O método foi dividido em quatro etapas: seleção da amostra de imagens, processo de avaliação, execução do experimento e a obtenção dos resultados compostos por imagem segmentada, grupo e a métrica J&F. Os resultados indicam que a técnica Kanezaki superou as demais. Pesquisas futuras são recomendadas para identificar se existe correlação entre as análises quantitativa e qualitativa.

Palavras-chave:

DOI: 10.5151/siintec2020-IMAGESEGMENTATION

Referências bibliográficas

• [1] ZHU, W. et al. Neurreg: Neural registration and its application to image segmentation. In:The IEEE Winter Conference on Applications of Computer Vision. [S.l.: s.n.], 2020. p. 3617–3626 2 AGRAWAL, S.; NATU, P. Segmentation of moving objects using numerous background subtraction methods for surveillance applications. International Journal of Innovative Technology and Exploring Engineering(IJITEE), v. 9, n. 3, p. 2553–2563, 2020. 3 GONZALES, R. C.; WOODS, R. E. Digital image processing. [S.l.]:Prentice hall New Jersey, 2002. 4 Ang, J. C. et al. Supervised, unsupervised, and semi-supervised feature selection: A review on gene selection. IEEE/ACM Transactions onComputational Biology and Bioinformatics, v. 13, n. 5, p. 971–989, 2016. 5 JING, L.; TIAN, Y. Self-supervised visual feature learning with deep neural networks: A survey. IEEE Transactions on Pattern Analysis andMachine Intelligence, IEEE, 2020. 6 KHOREVA, A.; ROHRBACH, A.; SCHIELE, B. Video object segmentation with language referring expressions. In: SPRINGER. AsianConference on Computer Vision. [S.l.], 2018. p. 123–141. 7 WANG, Z.; WANG, E.; ZHU, Y. Image segmentation evaluation: a survey of methods. Artificial Intelligence Review, Springer, p. 1–38, 2020. 8 LIN, T.-Y. et al. Microsoft coco: Common objects in context. In:SPRINGER. European conference on computer vision. [S.l.], 2014. p.740–755. 9 YU, Z. et al. An adaptive unsupervised approach toward pixel clustering and color image segmentation. Pattern Recognition, Elsevier. 43, n. 5, p. 1889–1906, 2010. 10 TAN, K. S.; ISA, N. A. M.; LIM, W. H. Color image segmentation using adaptive unsupervised clustering approach. Applied Soft Computing, Elsevier, v. 13, n. 4, p. 2017–2036, 2013. 11 KANEZAKI, A. Unsupervised image segmentation by backpropagation.In: IEEE. 2018 IEEE international conference on acoustics, speech and signal processing (ICASSP). [S.l.], 2018. p. 1543–1547. 12 GAN, G.; MA, C.; WU, J. Data clustering: theory, algorithms, and applications. [S.l.]: Siam, 2007. 13 LOO, P. K.; TAN, C. L. Adaptive region growing color segmentation for text using irregular pyramid. In: SPRINGER. International Workshop On Document Analysis Systems. [S.l.], 2004. p. 264–275. 14 WALTERS-WILLIAMS, Janett; LI, Yan. Comparative study of distance functions for nearest neighbors. In: Advanced techniques in computing sciences and software engineering. Springer, Dordrecht, 2010. p. 79-84. 15 WESOLOWSKI, S.; DONY, R. D.; JERNIGAN, M. Global color image segmentation strategies: Euclidean distance vs. vector angle. In: IEEE. Neural Networks for Signal Processing IX: Proceedings of the 1999IEEE Signal Processing Society Workshop (Cat. No. 98TH8468). [S.l.], 1999. p. 419–428. 16 GALLEGO, G. et al. On the mahalanobis distance classification criterion for multidimensional normal distributions. IEEE Transactions onSignal Processing, IEEE, v. 61, n. 17, p. 4387–4396, 2013. 17 PERAZZI, F. et al. A benchmark dataset and evaluation methodology for video object segmentation. In: Computer Vision and PatternRecognition. [S.l.: s.n.], 2016

Como citar:

Santos, Lucas Lisboa dos; Pagano, Tiago ; Vacaro, Juliano ; Loureiro, Rafael ; Junior, Neilton ; Cunha, Guilherme da ; Nascimento, Erick Giovani Sperandio ; Ingrid WINKLER; "IMAGE SEGMENTATION FOR PEOPLE IDENTIFICATION: ANEVALUATION OF UNSUPERVISED TECHNIQUES", p-635-643. In: Anais do VI Simpósio Internacional de Inovação e Tecnologia. São Paulo: Blucher, 2020.
ISSN 23577592, DOI 10.5151/siintec2020-IMAGESEGMENTATION