<![CDATA[The home security system is one of the features that must be owned and considered for every homeowner who wants to have a home that is safe from theft and avoid other unwanted security disturbances. So we need a support system that is able to increase home security. In this study, the system created uses the face as security data. This system uses a webcam as a face image taker and is integrated with the Raspberry Pi. This system will apply the buzzer, LED, solenoid door lock and SIM800L modules as outputs of the system. This system uses HaarClassifier to detect faces, then Histogram of Oriented Gradient and k-Nearest Neighbor for face recognition. First the system will take the image captured by the webcam, then use face image detection with Haar-Classifier, then the facial image will be extracted using the HOG feature. After the face feature value is obtained, it will then be classified using the k-Nearest Neighbor algorithm. From the results of testing the accuracy of face detection is the best accuracy of 100% at a distance of 40cm. The results of the accuracy of face recognition at a distance of 40cm in total are equal to 87.5%. For testing the accuracy of integration between software and hardware produces an accuracy rate of 100%. The average time needed for the face recognition process is 13,28839 seconds.]]>
Copyrights © 2019
