Categorization of Silkworm Based on Chitin Glands using Image Processing

Abstract

This paper demonstrates a prototype for highly accurate identification of the silkworm pupa (Bombyx mori) gender using optical property. The methodology is to optical beam in the near infrared spectrum that can effectively and safely penetrate the body of a silkworm pupa. After the illumination, some of the basic operations of image processing like image thresholding, contour detection, blob filtering and image inversion processes are applied to remove the unwanted image noises and at the same time highlighted the gland that distinguishes the gender in silkworm. The proof of concept is experimentally done using three 633 nm wavelength Light emitting diodes (LED’s), a pi camera, and a computer. Some of the key features of this method include ease of implementation with cost reduction and high accuracy