Object classification has been done to various images. Animal classification has been done using segmentation and non-segmentation approach as initial stage. Context Aware Saliency (CAS) is a method that able to make the object area more dominant than the background in saliency mode so that it can be an alternative object segmentation process. The shape feature will taken based on saliency results using the Histogram of Oriented Gradient (HOG). The K-Nearest Neighbors (K-NN) used to classify mammal species based on HOG features from saliency images. The dataset used in this study is LHI-Animal-Faces. The results obtained show that animal species that can be recognized well are cats and tigers, while sheep, dogs, and pigs have not been able to be recognized properly.
Copyrights © 2019