在本篇論文中,我們提出了一個結合智慧型手機拍照和手部辨識方法的應用程式。我們主要的目的是能利用手部辨識,提高手機拍照時的便利性及讓使用者拍照體驗好感提升。 在本論文中,我們會介紹如何辨識手部動作並結合於手機拍照。首先,透過手機鏡頭擷取影像資訊後,將影像壓縮至適合大小,並將影像由RGB轉成YCbCr色彩空間,之後利用適合的閥值選取出膚色範圍作二值化處理;其次,根據我們找出的膚色範圍作斷開處理,以去除背景雜訊,將大範圍的臉部及手部區別出來,我們預設手部面積比臉部面積大,故選取面積較小的手部膚色區域為我們需要的判斷對象;第三,根據我們所選擇的手部區域作邊界偵測,利用輪廓判別手部動作,如判定為我們所設定之手勢,則儲存該影像於手機中;最後,我們會呈現出我們的實驗結果。 在本篇論文中,我們的主要貢獻有以下幾點: (1) 提供一個全新的拍照體驗: 有別於延遲拍攝及微笑偵測,利用手勢作為照相快門決定的時機,讓拍照能更輕鬆、更直覺。 (2) 提高影像辨識速度: 利用簡單快速的辨識方法,加快手機運算速度,減少拍照時產生的延遲。 (3) 系統的拓展性: 將辨識系統寫成手機APP程式,讓使用者可自行更新,不需要花費大筆金錢作硬體更新,即可享有方便的拍照軟體。
In this thesis, we propose an application to hand gesture detection on smartphone. The main propose in this thesis is using hand gesture detection to increase convenience and users’ experience of taking photo on smartphone. In this thesis, we will introduce how to use hand gesture detection on smartphone’s camera. First, in order to find out the skin color, we capture image through smartphone’s camera and compress it to fit size we need. Then we process the image into binary image by converting it from RGB to YCbCr in an error rate for the average value let it be our skin color threshold. Second, according to the binary image, we distinguish a wide region of face and hand from background by opening of the binary image. We choose the small area of skin color region, which we suppose that skin color region of hand is smaller than face. Third, focusing on the hand region, we use Canny filter to find the edge of hand gesture, after that, we identify the gesture and store the image into smart phone if the gesture is what we set previously. Finally, we will present the simulation experiments and the comparison with the other researches. The contributions of our research are as follows: (1) Provide a new experience of taking photo: Unlike delay shooting and smile detection, it is easier and more direct by using hand gesture to decide the moment of taking photo. (2) Increase the speed of image processing: Use simple identification method to reduce delay created by operation of smartphone camera. (3) Expandable of system: We design this method in smartphone application. User can update software by themselves in free.