| Autor: |
Quoc-Loc Duong, Xuan-Vinh Nguyen, The-Manh Nguyen, Trong-Hop Do, Gia-Huy Lam, Nhu-Ngoc Dao, Bao-Long Le, Quang D. Tran, Quang-Nhat Le |
| Rok vydání: |
2021 |
| Předmět: |
|
| Zdroj: |
2021 IEEE International Conference on Communication, Networks and Satellite (COMNETSAT). |
| Popis: |
Video streaming is a multimedia service that continuously transmits the data over the Internet and presents the content on user screens without predownloading the entire video. Augmented video streaming is an advanced version of video streaming, where the video is enriched with additional embedded information in video frames. These additional data aim to provide a better user experience. Using QR code is one of the efficient approaches to incorporate information into video streams in this context. However, receiving the data in the embedded QR code is considered a challenging task owing to video quality and view angles. This paper proposes a lightweight two-stage QR code decoder for augmented video streaming using deep learning technologies. In the first stage, the position of the embedded QR code is detected using an online object detection algorithm. In the second stage, the detected region of the QR code is fed into a QR code reader to extract the embedded data. The experimental results show that the proposed decoder achieves high performances in terms of response time and decoding accuracy while being very lightweight, which is promising to be implemented in smartphones. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
