近年來由於行動網路的快速發展,讓行動裝置有了爆發性的成長,幾乎人手一支智慧型手機,大部份手持行動裝置也都把相機功能列為基本配備,所以藉由相機便可以來實現不同的影像識別與影像處理如QR Code等。本論文主要是研究如何在Android上利用計算機視覺處理函式庫OpenCV4Android (Open Compute Vision)來開發一個影像辨識相關的應用程式,應用程式透過手持行動裝置所內建的相機來擷取影像,並利用OpenCV4Android所提供強大的計算機視覺運算、處理的函式庫,讓我們可以透過OpenCV4Android處理所擷取的影像,接著我們會利用OpenCV所提供的影像辨識函式,分析與判斷所擷取到的影像是屬於哪一類電器產品,再透過遠端控制模組,進一步直接利用行動裝置對電器進行遠端操作。
The fast development in mobile network has explosive growth in recent years, smart phone becomes a must-have device for everyone. In particular, camera becomes a basic feature in most mobile devices. As a result, utilizing camera to perform image recognition, image input and image processing are accessible to everyone in such scenario. This research focuses on the design and implementation on an application to perform appliance recognition based on the OpenCV4Android (Open Compute Vision) library on the Android platform. This application captured an image with the build-in camera in a handheld mobile device, which enabled us to further process the screenshots with the powerful computer vision library of OpenCV4Android. With the use of OpenCV library, we are able to analyze and evaluate the screenshot and identify the appliance. In addition, once the appliance can be identified, we are able to control the appliance. The results show that our system performs well and this work can be further integrated in the area of green technology and smart home.