Silkworm eggs detection and classification using image analysis

Machine vision has been applied to various material inspection processes in agricultural industry in order to achieve fast and accurate operation. In this paper, we propose an image processing technique for silkworm eggs analysis. The proposed technique is able to identify individual egg objects and their types. The proposed technique was implemented and developed into software interface. The software was installed at the Thailand Sericulture center and it is going to be used in the real practical scenario. To evaluate our proposed technique, the approach was evaluated on 60 sample images from 7 shade groups of silkworm's egg sample. The obtained accuracies are nearly to 90% for various kinds of eggs types in both detection and classification. The results of this study are useful and they are served as an important base for future development of practical quality control technique.