Notícias

Banca de DEFESA: TIAGO DAVI OLIVEIRA DE ARAUJO

Uma banca de DEFESA de DOUTORADO foi cadastrada pelo programa.
DISCENTE: TIAGO DAVI OLIVEIRA DE ARAUJO
DATA: 02/02/2022
HORA: 11:00
LOCAL: https://videoconf-colibri.zoom.us/j/86315899300
TÍTULO:

A model for automated support for recognition, extraction, customization and reconstruction of static charts


PALAVRAS-CHAVES:

chart recognition, chart reconstruction, deep learning, information visualization, perspective correction, augmented reality, real-time visualization, interaction


PÁGINAS: 105
GRANDE ÁREA: Ciências Exatas e da Terra
ÁREA: Ciência da Computação
SUBÁREA: Metodologia e Técnicas da Computação
ESPECIALIDADE: Processamento Gráfico (Graphics)
RESUMO:

Data charts are widely used in our daily lives, being present in regular media, such as newspapers, magazines, web pages, books, and many others. A well-constructed data chart leads to an intuitive understanding of its underlying data and in the same way, when data charts have wrong design choices, a redesign of these representations might be needed. However, in most cases, these charts are shown as a static image, which means that the original data are not usually available. Therefore, automatic methods could be applied to extract the underlying data from the chart images to allow these changes. The task of recognizing charts and extracting data from them is complex, largely due to the variety of chart types and their visual characteristics. Computer Vision techniques for image classification and object detection are widely used for the problem of recognizing charts, but only in images without any disturbance. Other features in real-world images that can make this task difficult are not present in most literature works, like photo distortions, noise, alignment, etc. Two computer vision techniques that can assist this task and have been little explored in this context are perspective detection and correction. These methods transform a distorted and noisy chart in a clear chart, with its type ready for data extraction or other uses. The task of reconstructing data is straightforward, as long the data is available the visualization can be reconstructed, but the scenario of reconstructing it on the same context is complex. Using a Visualization Grammar for this scenario is a key component, as these grammars usually have extensions for interaction, chart layers, and multiple views without requiring extra development effort. This work presents a model for automated support for custom recognition, and reconstruction of charts in images. The model automatically performs the process steps, such as reverse engineering, turning a static chart back into its data table for later reconstruction, while allowing the user to make modifications in case of uncertainties. This work also features a model-based architecture along with prototypes for various use cases. Validation is performed step by step, with methods inspired by the literature. This work features three use cases providing proof of concept and validation of the model. The first use case features usage of chart recognition methods focused on documents in the real-world, the second use case focus on vocalization of charts, using a visualization grammar to reconstruct a chart in audio format, and the third use case presents an Augmented Reality application that recognizes and reconstructs charts in the same context (a piece of paper) overlaying the new chart and interaction widgets. The results showed that with slight changes, chart recognition and reconstruction methods are now ready for real-world charts, when taking time, accuracy and precision into consideration.


MEMBROS DA BANCA:
Presidente - 2325270 - BIANCHI SERIQUE MEIGUINS
Externo à Instituição - DANIEL JORGE VIEGAS GONÇALVES
Externo à Instituição - JOAQUIM JOÃO ESTRELA RIBEIRO SILVESTRE MADEIRA
Externo à Instituição - JOSÉ GUSTAVO DE SOUZA PAIVA
Externo à Instituição - MARCELO DE PAIVA GUIMARÃES
Externo à Instituição - MARIA BEATRIZ ALVES DE SOUSA SANTOS
Externo à Instituição - MARIA BEATRIZ DUARTE PEREIRA DO CARMO
Externo à Instituição - ÓSCAR EMANUEL CHAVES MEALHA
Notícia cadastrada em: 01/02/2022 19:06
SIGAA | Centro de Tecnologia da Informação e Comunicação (CTIC) - (91)3201-7793 | Copyright © 2006-2024 - UFPA - bacaba.ufpa.br.bacaba1