On-line identification of silkworm pupae gender by short-wavelength near infrared spectroscopy and pattern recognition technology

Abstract

The gender identification of silkworm pupae is a critical step in the sericulture industry's breeding process. In this study, a low cost, short-wavelength (815-1075 nm) near infrared (NIR) spectrometer combined with multivariate spectra evaluation methods was used to establish calibration models for the on-line identification of female and male pupae of eight silkworm varieties (Hibiscus, Jingsong, 932, Xiang Hui, 7532×Xiang Hui, Haoyue B, Jingsong B, and 7532). The diffuse reflection short-wavelength spectra were recorded, and then principal component analysis (PCA), linear discriminant analysis (LDA), and partial least squares discriminant analysis (PLSDA) were tested for calibration model development. The PCA and LDA results showed, that spectral differences between the female and male silkworm pupae existed, however, the two evaluation techniques could not separate the female and male silkworm pupae with the required accuracy. The PLSDA calibration models, on the other hand, could separate the pupae according to their gender with the necessary prediction accuracy of >98.44%. Thus, it has been proved, that a low-cost, short-wavelength range NIR spectrometer in combination with a PLSDA calibration routine can be successfully applied for the reliable on-line identification of female and male silkworm pupae.

Integration of Artistic Visual Elements in Packaging Design Based upon Pattern Recognition Technology

“…e average image of all the learned templates is the displayed template. When the number of learned template images reaches the specified number, the learning is stopped, and the learning process of the template images ends at this time [19,20].…”

Intelligent identification of segmental line losses in low-voltage networks under graphical consistency line variation relationship analysis

“…Line loss identification work as an important link in line loss management, can effectively guarantee the normal operation of the power grid lines. By effectively identifying the area of line loss in the power grid, the aging line can be replaced or repaired in a timely manner, thus improving the production efficiency of power grid operation and providing protection for the good operation of power supply enterprises [1]. At this stage, there are mainly two means for the identification of power grid line loss, one is the conventional identification method based on manual determination; the other is the intelligent identification method based on mathematical model.…”