Proximal Sensing and Data Analysis for Agricultural Products
A.Y. 2023/2024
Learning objectives
Knowledge of the theory of the main agricultural products optical properties and of non-destructive analysis techniques. Knowledge of the main categories of optical instruments available on the market. Knowledge of the main multivariate data processing techniques. Provide skills for a practical use of the instrumentation and for data interpretation to better manage the pre- and post-harvest phases.
Expected learning outcomes
Acquire skills to use benchtop and portable optical instrumentation. Correctly acquire optical data on agricultural samples. Acquire skills to extract useful information from multivariate data and to calculate classification and quantitative predictive models. Perform multivariate statistical analysis using specific software. Manage modeling output to undertake decisions and for the correct management of production.
Lesson period: First semester
Assessment methods: Esame
Assessment result: voto verbalizzato in trentesimi
Single course
This course can be attended as a single course.
Course syllabus and organization
Single session
Responsible
Lesson period
First semester
Course syllabus
1. Interaction of the electromagnetic radiation with matter. Optical properties of agricultural products. Vibrational spectroscopy. Absorption bands of the spectrum in the visible and near infrared range (vis/NIR).
2. Instrumentation analysis (vis/NIR spectroscopy, image analysis and hyperspectral imaging, thermal imaging). Full spectral range and simplified instruments with specific spectral bands. Bench top and portable instruments.
3. Experimental setup for optical measurements in lab scale or open field environment. Fundaments of experimental design. Management of environmental interference on the optical data.
4. Spectra interpretation. Data pretreatments. Exploratory analysis of multivariate data: principal components analysis (PCA). Data modeling: qualitative classification models and quantitative predictive models. Model calibration and validation. Practical exercises of data analysis using specific software.
5. Examples of proximal sensing applications in preharvest for the main horticultural chains (e.g. measurement of yield, quality parameters, ripeness degree, optimal harvest time estimation, early diseases detection, monitoring of plant stress). Examples of proximal sensing applications in postharvest (e.g. measurement of quality parameters, degree of senescence, postharvest life). Analysis of real case studies and experimental measurements.
2. Instrumentation analysis (vis/NIR spectroscopy, image analysis and hyperspectral imaging, thermal imaging). Full spectral range and simplified instruments with specific spectral bands. Bench top and portable instruments.
3. Experimental setup for optical measurements in lab scale or open field environment. Fundaments of experimental design. Management of environmental interference on the optical data.
4. Spectra interpretation. Data pretreatments. Exploratory analysis of multivariate data: principal components analysis (PCA). Data modeling: qualitative classification models and quantitative predictive models. Model calibration and validation. Practical exercises of data analysis using specific software.
5. Examples of proximal sensing applications in preharvest for the main horticultural chains (e.g. measurement of yield, quality parameters, ripeness degree, optimal harvest time estimation, early diseases detection, monitoring of plant stress). Examples of proximal sensing applications in postharvest (e.g. measurement of quality parameters, degree of senescence, postharvest life). Analysis of real case studies and experimental measurements.
Prerequisites for admission
The course requires basic physical and mathematical knowledge to face the crucial aspects of the theory and of the instrumentation. Basic statistical knowledge is important to address the aspects of multivariate data processing.
Teaching methods
The course includes 5 CFU of frontal teaching and 1 CFU with practical use of the instrumentation and numerical simulations using specific statistical software. During the lectures theoretical lessons, practical use of instrumentation and numerical simulations are presented to facilitate the acquisition of the expected expertise through practical case studies.
Teaching Resources
1. Slides and Lecture Notes, videorecording of the lessons.
2. Donald A. Burns (Editor), Emil W. Ciurczak (Editor). Handbook of Near-Infrared Analysis (Practical Spectroscopy). CRC Press.
3. Esbensen K.H. (2006). Multivariate data analysis - in practice. Aalborg University, Esbjerg.
4. Guidetti, R., Beghi, R., & Giovenzana, V. (2012). Chemometrics in food technology. Chemometrics in practical applications, 217-252.
2. Donald A. Burns (Editor), Emil W. Ciurczak (Editor). Handbook of Near-Infrared Analysis (Practical Spectroscopy). CRC Press.
3. Esbensen K.H. (2006). Multivariate data analysis - in practice. Aalborg University, Esbjerg.
4. Guidetti, R., Beghi, R., & Giovenzana, V. (2012). Chemometrics in food technology. Chemometrics in practical applications, 217-252.
Assessment methods and Criteria
The final exam will be an unique oral exam: theoretical "open" questions (3 questions) will be proposed based on the topic presented during the course. No intermediate tests will be planned.
AGR/09 - AGRICULTURAL MACHINERY AND MECHANIZATION - University credits: 6
Practicals: 16 hours
Lessons: 40 hours
Lessons: 40 hours
Professor:
Beghi Roberto
Educational website(s)
Professor(s)
Reception:
by appointment only
Department of Agricultural and Environmental Sciences - via Celoria 2, Milano