Chartem: Reviving Chart Images with Data Embedding

Jiayun Fu,Bin Zhu,Weiwei Cui,Song Ge,Yun Wang,Haidong Zhang,He Huang,Yuanyuan Tang,Dongmei Zhang,Xiaojing Ma

DOI: https://doi.org/10.1109/tvcg.2020.3030351

IF: 5.2

2021-02-01

IEEE Transactions on Visualization and Computer Graphics

Abstract:In practice, charts are widely stored as bitmap images. Although easily consumed by humans, they are not convenient for other uses. For example, changing the chart style or type or a data value in a chart image practically requires creating a completely new chart, which is often a time-consuming and error-prone process. To assist these tasks, many approaches have been proposed to automatically extract information from chart images with computer vision and machine learning techniques. Although they have achieved promising preliminary results, there are still a lot of challenges to overcome in terms of robustness and accuracy. In this paper, we propose a novel alternative approach called Chartem to address this issue directly from the root. Specifically, we design a data-embedding schema to encode a significant amount of information into the background of a chart image without interfering human perception of the chart. The embedded information, when extracted from the image, can enable a variety of visualization applications to reuse or repurpose chart images. To evaluate the effectiveness of Chartem, we conduct a user study and performance experiments on Chartem embedding and extraction algorithms. We further present several prototype applications to demonstrate the utility of Chartem.

computer science, software engineering

What problem does this paper attempt to address?

This paper presents a solution to the problem of embedding chart information into images for reusability and modification. Currently, charts are often stored as bitmap images, which requires recreating the entire chart when changing styles or data, resulting in time-consuming and error-prone tasks. To address this problem, the paper proposes a method called Chartem. Chartem designs a data embedding scheme that encodes a large amount of information into the background of chart images without interfering with human perception of the chart. The embedded information can be extracted from the image, enabling various visualization applications to reuse or repurpose chart images. The effectiveness of Chartem is evaluated through user studies and performance experiments, and several prototype applications are demonstrated to prove its practicality. This approach differs from computer vision and machine learning-based methods for automatic chart information extraction, as it directly addresses the problem at its source, ensuring information integrity and accuracy.