PEMANFAATAN METODE K-NEAREST NEIGHBOR PADA KLASIFIKASI IMAGE BERDASARKAN POLA FITUR DAN TEKSTUR
Abstrak: Content-based image
search can use Content Based Image Retrieval (CBIR). CBIR works by measuring
the similarity of query images with all the images in the database so that the
query cost is directly proportional to the number of images in the database.
Limiting the range of image search by way of classification is one way to
reduce the query cost on CBIR. Application of K-Nearest Neighbor method aims to
classify the image as well as to measure the level of accuracy and time of
classification. In this study built a software that can extract the color and
texture features of an image by using the Color Histogram method and the Edge
Histogram Descriptor. The results of the feature extraction process are then
used by the software in the learning process and classification by the
K-Nearest Neighbor method. The software is built with structured analysis and
design methods then implemented using VB.net programming language The final
result of classification is then tested with parameter level accuracy and
classification time. The test results show that the combination of color and
texture features provides a higher level of accuracy than classification based
on features and textures but requires longer classification time.
Penulis: Nurul Fuad
Kode Jurnal: jptinformatikadd170363
