Tape-Shaped, Multiscale, and Continuous-Readable Fiducial Marker for Indoor Navigation and Localization Systems.

The present study proposes a fiducial marker for location systems that uses computer vision. The marker employs a set of tape-shaped markers that facilitate their positioning in the environment, allowing continuous reading to cover the entire perimeter of the environment and making it possible to minimize interruptions in the location service. Because the marker is present throughout the perimeter of the environment, it presents hierarchical coding patterns that allow it to be robust against multiple detection scales. We implemented an application to help the user generate the markers with a floor plan image. We conducted two types of tests, one in a 3D simulation environment and one in a real-life environment with a smartphone. The tests made it possible to measure the performance of the tape-shaped marker with readings at multiple distances compared to ArUco, QRCode, and STag with detections at distances of 10 to 0.5 m. The localization tests in the 3D environment analyzed the time of marker detection during the journey from one room to another in positioning conditions (A) with the markers positioned at the baseboard of the wall, (B) with the markers positioned at camera height, and (C) with the marker positioned on the floor. The localization tests in real conditions allowed us to measure the time of detections in favorable conditions of detections, demonstrating that the tape-shaped-marker-detection algorithm is not yet robust against blurring but is robust against lighting variations, difficult angle displays, and partial occlusions. In both test environments, the marker allowed for detection at multiple scales, confirming its functionality.

Pactolo Bar: An Approach to Mitigate the Midas Touch Problem in Non-Conventional Interaction.

New ways of interacting with computers is driving research, which is motivated mainly by the different types of user profiles. Referred to as non-conventional interactions, these are found with the use of hands, voice, head, mouth, and feet, etc. and these interactions occur in scenarios where the use of mouse and keyboard would be difficult. A constant challenge in the adoption of new forms of interaction, based on the movement of pointers and the selection of interface components, is the Midas Touch (MT) problem, defined as the involuntary action of selection by the user when interacting with the computer system, causing unwanted actions and harming the user experience during the usage process. Thus, this article aims to mitigate the TM problem in interaction with web pages using a solution centered on the Head Tracking (HT) technique. For this purpose, a component in the form of a Bar was developed and inserted on the left side of the web page, called the Pactolo Bar (PB), in order to enable or disable the clicking event during the interaction process. As a way of analyzing the effectiveness of PB in relation to TM, two stages of tests were carried out based on the collaboration of voluntary participants. The first step aims to find the data that would lead to the best configuration of the BP, while the second step aims to carry out a comparative analysis between the PB solution and the eViacam software, whose use is also focused on the HT technique. The results obtained from the use of PB were considered promising, since the analysis of quantitative data points to a significant prevention of involuntary clicks in the iteration interface and the analysis of qualitative data showed the development of a better user experience due to the ease of use, which can be noticed in elements such as the PB size, the triggering mechanism, and its positioning in the graphical interface. This study benefits in the context of the user experience, because, when using non-conventional interactions, basic items such as aspects of the graphic elements, and interaction events raise new studies that seek to mitigate the problem of the Midas Touch.

A Real-World Approach on the Problem of Chart Recognition Using Classification, Detection and Perspective Correction.

Data charts are widely used in our daily lives, being present in regular media, such as newspapers, magazines, web pages, books, and many others. In general, a well-constructed data chart leads to an intuitive understanding of its underlying data. 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. Other features in real-world images that can make this task difficult are 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. This paper proposes a classification, detection, and perspective correction process that is suitable for real-world usage, when considering the data used for training a state-of-the-art model for the extraction of a chart in real-world photography. The results showed that, with slight changes, chart recognition methods are now ready for real-world charts, when taking time and accuracy into consideration.

A Model to Support Fluid Transitions between Environments for Mobile Augmented Reality Applications.

The adaptability between different environments remains a challenge for Mobile Augmented Reality (MAR). If not done seamlessly, such transitions may cause discontinuities in navigation, consequently disorienting users and undermining the acceptance of this technology. The transition between environments is hard because there are currently no localization techniques that work well in any place: sensor-based applications can be harmed by obstacles that hamper sensor communication (e.g., GPS) and by infrastructure limitations (e.g., Wi-Fi), and image-based applications can be affected by lighting conditions that impair computer vision techniques. Hence, this paper presents an adaptive model to perform transitions between different types of environments for MAR applications. The model has a hybrid approach, choosing the best combination of long-range sensors, short-range sensors, and computer vision techniques to perform fluid transitions between environments that mitigate problems in location, orientation, and registration. To assess the model, we developed a MAR application and conducted a navigation test with volunteers to validate transitions between outdoor and indoor environments, followed by a short interview. The results show that the transitions were well succeeded, since the application self-adapted to the studied environments, seamlessly changing sensors when needed.