In this paper, the near-infrared spectral data of five different types of starch were collected, and the starch species identification model was constructed by using a quaternion convolutional neural network (QCNN), we proved that the qualitative model based on QCNN has obtained higher prediction accuracy than traditional qualitative models. In the experimental results, the classification accuracy of QCNN for five different starches reached 0.996. The results show that the combination of the quaternion spectral fusion method and deep learning is more conducive to extracting and mining the deep information of NIR spectra and has important research significance and application value in the field of near-infrared spectroscopy technology
KEYWORDS: Scientific research, Charge-coupled devices, Optoelectronics, Electronics, Video processing, Video, Signal processing, New and emerging technologies
The CCD principle and application course is professional and comprehensive. It involves many subject contents. The course content includes eight aspects. In order to complete the teaching tasks within a limited time, improve the classroom teaching quality and prompt students master the course content faster and better, so the multidimensional interactive classroom teaching is proposed. In the teaching practice, the interactive relationship between the frontier science, scientific research project, living example and classroom content is researched detailedly. Finally, it has been proved practically that the proposed multidimensional interactive classroom teaching can achieved good teaching effect.
Quantitative analysis of the simulated complex oil spills was researched based on PSO-LS-SVR method. Forty simulated mixture oil spills samples were made with different concentration proportions of gasoline, diesel and kerosene oil, and their near infrared spectra were collected. The parameters of least squares support vector machine were optimized by particle swarm optimization algorithm. The optimal concentration quantitative models of three-component oil spills were established. The best regularization parameter C and kernel parameter σ of gasoline, diesel and kerosene model were 48.1418 and 0.1067, 53.2820 and 0.1095, 59.1689 and 0.1000 respectively. The decision coefficient R2 of the prediction model were 0.9983, 0.9907 and 0.9942 respectively. RMSEP values were 0.0753, 0.1539 and 0.0789 respectively. For gasoline, diesel fuel and kerosene oil models, the mean value and variance value of predict absolute error were -0.0176±0.0636 μL/mL, -0.0084±0.1941 μL/mL, and 0.00338±0.0726 μL/mL respectively. The results showed that each component’s concentration of the oil spills samples could be detected by the NIR technology combined with PSO-LS-SVR regression method, the predict results were accurate and reliable, thus this method can provide effective means for the quantitative detection and analysis of complex marine oil spills.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.