Currently, many consumers electronic are utilizing personal identification technique such as ID, fingerprint, face, Iris and palmprint recognition for security reason. Among other biometric traits, face image is easy to capture using Pi Camera. This paper proposes facial recognition using principle component analysis (PCA) method implemented using Raspberry Pi embedded processor. The algorithm is model using PC based method to identify the best eigenvector to represent facial image. The pre-calculated statistical parameter is then used to implement the PCA algorithm in Raspberry Pi embedded processor. OpenCV image processing libraries is use to support the basic task of image pre-processing such as cropping, resize and color conversion. In the propose work, the whole system is implemented in low cost processor to evaluate the performance in terms of recognition rates and processing time. The proposed method includes three main phases. The pre-processing phase processes facial ORL image and to collect significant information. The second phase extracts important features from image and decreases the size of the image using the PCA method. This phase uses a linear projection technique to decrease redundancies and remove noise from the image. Moreover, this strategy also improves the strength of discrimination power in the feature space. The Euclidean distance classifier is utilized in the third phase of classification. The best recognition rates achieved using the propose method is 82.5% implemented using low cost embedded system.
