螢幕-相機通訊系統為現今最為常見的一種短距離通訊系統,空間 域編碼方式的螢幕-相機通訊系統目前已被廣泛使用於生活當中。然而 使用空間域編碼方式的螢幕-相機通訊系統也面臨了一些挑戰,包括通 訊距離的不足、容易受到影像模糊的影響和資料傳輸量上的限制,導 致空間域編碼方式的螢幕-相機通訊系統無法適用於更多的應用場景。 為了改善螢幕-相機通訊系統之通訊距離不足問題,我們選擇較不容易 受到影像模糊影響之頻率域編碼方式來設計傳輸端呈現的二維碼。本 篇論文中三個主要的挑戰為螢幕-相機間亮度的非線性關係、二維碼的 透視變形和影像模糊的問題。針對這三個挑戰,我們分別提出亮度校 正、重取樣與通道估測的方法來降低解碼的錯誤率。為探討這些方法 的效果,我們採用三種大小不同的螢幕及兩種相機來進行測試。實驗 結果顯示,亮度校正能有效的降低至多 20% 的通訊錯誤率。而重採樣 的方法則能有效的降低圖形形變造成的解碼錯誤。
Screen-to-camera communication is one of the most popular short-range communication systems in recent years. Two main encoding schemes used in screen-to-camera communications are spatial encoding and frequency encoding scheme. Even though spatial encoding scheme, such as QR code, has been widely deployed in our daily life, spatial encoding scheme suffers from short communication range, and is vulnerable to the blurring effect. On the other hand, frequency encoding scheme is considered to be more resilient to the blurring effect. As our ultimate goal is to increase the communication range of screen-to-camera communications, frequency encoding scheme is adopted in this work to generate the transmitted code. There are three major challenges in this work: nonlinear relationship between the transmitted intensity by the screen and received intensity by the camera, perspective distortion and the blurring effect. To address these challenges, we proposed nonlinearity calibration, resampling and channel estimation respectively to reduce the decoding error. To evaluate our proposed methods, three kinds of screens with different sizes and two cameras are used to carry out the experiments. The experimental results show that nonlinearity calibration is able to reduce the error rate with an error rate drop of at most 20%. In addition, the resampling algorithm can effectively mitigate the error caused by the distortion.