Bottleneck RGB Features for Tea Clones Identification

Abstract

As each tea clone may produce different quality of tea, it is important to have them identified in the field. Tea Clones identification is one application of ICT technologies in agriculture. Tea clones may have very similar characteristics between them, required to have a good amount of data to train a machine learning-based classifiers to have good performances. However, we may have to deal with a small amount of data in many cases. To overcome this, we propose to use an encoder-based feature reduction to produce RGB-based bottleneck features. The output features are then fed into an SVM classifier. We evaluate our features on the classification of two tea clones of the Gambung Assamica (GMB) series. Our experimental results show that our proposed features achieve better performance than using full dimensions RGB.