SISTEM PENGENALAN WAJAH TAMPAK SAMPING DENGAN METODE HARRIS CORNER DAN BACKPROPAGATION
Daftar Isi:
- Face recognition systems generally use front view face objects . In this research, the side view face was used as the main object in the side-view facial recognition system. This system was developed by using two methods, the detection system with Harris corner and classification for side view face recognition using backpropagation. Harris corner will show the corner of the side structure of the side view, this corner is then used as feature extraction. To get the vertex, it is necessary to pay attention to the parameters k, sigma, and threshold. After that the point that has been obtained is calculated Euclidean distance between the corners to be used as an input vector in the next method, namely backpropagation. The number of input vectors is the most number of corner obtained from all sidelooking facial images. The introduction is done by using backpropagation by comparing the output value of the training data that has been obtained in the training process compared to the image value tested. In this study testing of data that has been trained and data that have not been trained, the data that has been tested for testing 100 image samples with 20 target names results in an introduction with a percentage of sensitivity of 72.25%, specificity of 98.72, and accuracy of 97, 5%. While testing for data that has not been trained as much as 40 test data samples with 20 target names get a percentage of sensitivity 61.17%, specificity 98.18%, and accuracy of 95.38%