Machine vision and AI for Agri&Food applications


Nuzzi Cristina Nuzzi

Cristina Nuzzi

University of Brescia, Italy

Pasinetti Simone Pasinetti

Simone Pasinetti

University of Brescia, Italy

Gregorio López Eduard Gregorio López

Eduard Gregorio López

University of Lleida, Spain

Gené-Mola Jordi Gené-Mola

Jordi Gené-Mola

Institute of Agrifood Research and Technology, Spain


The Agri&Food sector is facing a revolution due to the increasing demand of agricultural products for the food, energy, and manufacturing sectors. As a consequence, technologies traditionally adopted by industry are now being adopted in agriculture with the additional constraints of in-field adverse conditions such as meteorological agents (rain, snow, wind, humidity, etc.), natural light interference, non-optimal temperatures, absence of electrical current and of internet connection in the whole field.

One of the most promising technologies that is intensively being used to advance the research field is vision, because a large amount of information can be acquired and processed to extract meaningful data and perform measurements, from mechanical dimensions and quantities of fruits and leaves to the movement patterns of insects. Vision-based technology can acquire data either in 2D or 3D according to the needs, allowing researchers to obtain 3D data of plants, fruits and even the whole orchard to estimate their volumes and shapes. Vision is also the preferred medium to empower robotic agents and automated machines for in-field fruit picking, slicing or monitoring in general. Furthermore, machine vision is not limited to the visible spectrum: infrared and UV wavelengths also provide valuable data that can be used to monitor and analyze biological and chemical phenomena in plants, food, and livestock.

Thanks to the advances of artificial intelligence (which includes machine learning and deep learning methods), a plethora of models and algorithms are available to the research community to interpret and analyze complex data such as images, image cubes or even 3D point clouds. Thus, this session's objective is to explore and discuss the potential of new AI-based methods to analyze vision data, either 2D and 3D, to perform measurements on plants, forests, fruits, orchards, insects, livestock and food in general.

Special attention will be given to applications leveraging technology that acquires visual data beyond the visible spectrum, such as thermal cameras, NIR, UV and UV fluorescence, multispectral and hyperspectral imaging. Finally, the statistical and metrological validation of the proposed approach are fundamental aspects to be discussed for the works presented in this session.


We welcome contributions that cover the following topics:

  • image processing and computer vision for measuring quantities;
  • beyond visible technology: thermal vision, NIR, MIR, UV, multispectral, hyperspectral;
  • applications leveraging 3D vision;
  • geometrical measurements of plants, fruits, insects and livestock;
  • chemical residuals detection;
  • plant transpiration leveraging thermal vision;
  • phenotyping leveraging image processing;
  • fruit counting and sizing;
  • image vision to empower agricultural robotic agents;
  • metrological validation;
  • AI for data processing.


Dr. Cristina Nuzzi, (born 1993) has been an assistant professor (Italian RTD-A) in Mechanical and Thermal Measurements (ING/IND-12) since 2022. She holds both a bachelor's and a master's degree in automation engineering, received in 2015 and 2017, respectively, from the University of Brescia. She completed her Ph.D. in Mechanical and Industrial Engineering (track Applied Mechanics) in 2020 at the University of Brescia with a dissertation about the concept of "meta-collaborative workstations" and software designed to communicate with robots developed using vision systems and deep learning models.
She has been a member of the "Vision systems for mechatronics and agriculture" and "Mechanical and thermal measurements for human wellbeing and industry" research groups (https://www.linkedin.com/company/mmt-lab/) since her Ph.D. years.
Despite her young age, she is the first author of several publications centered around measurements in unstructured environments, mostly leveraging contactless or wearable sensors, and data analysis. Her topics of interest include measurements for smart agriculture, plant diseases and pest monitoring, human motion and biomechanics, and collaborative robots.
Dr. Nuzzi has wide experience in the development, training, and utilization of deep learning models for vision data and image processing techniques. Her expertise also includes data management and dataset creation, 3D data elaboration and processing, and software deployment for target embedded hardware. She co-tutored several Master's degree theses on the topic of measurements and intelligent algorithms.

Dr. Simone Pasinetti (born 1985) received his B.S. degree and M.S. degree (cum laude) in Automation Engineering from the University of Brescia in 2009 and 2011 respectively. Since 2011 he works in the Mechanical and Thermal Measurement Laboratory (MMTLAB) at University of Brescia. Since 2020 he is an Assistant Professor in Mechanical and Thermal Measurements (ING-IND/12) at the University of Brescia, Italy. He has received the national scientific qualification in May 2021. He is currently the Head of the Vision Systems for Mechatronic and Agriculture division of the MMTLAB. He has a strong background in developing and characterizing vision systems and algorithms based both on deterministic and advanced (i.e. machine and deep learning vision processing) machine vision. During his research career, he worked mainly on the development of measurement systems based on 2D and 3D vision. In particular, he worked on human kinematic analysis, liquid lens objectives applications, and 3D vision systems development. He studied new measurement techniques based on vision and developed new systems for applications for different fields such as biomechanics, medicine, robotics, cultural heritage, and industry. In recent years, his main research interests include vision and measurement systems for the agriculture and food industries.

Dr. Eduard Gregorio is an associate professor at the Department of Agricultural and Forest Sciences and Engineering of the University of Lleida (Spain) and member of the Research Group in AgroICT & Precision Agriculture (https://www.grap.udl.cat/en/). He holds a BS degree in mechanical engineering from the University of Lleida and a MS degree in industrial engineering from the Technical University of Catalonia (Spain). His scientific interests include the application of sensors in agriculture, LiDAR remote sensing, and precision agriculture. He has led the development of an innovative range‐resolved LiDAR system for real‐time monitoring of pesticide spray drift. His recent research focuses on the application of low-cost photonic sensors for the geometric characterization of crops and for fruit detection and sizing.

Jordi Gené-Mola holds a B.S. degree in Mechanical Engineering (UdL, 2013), an M.Sc. degree in Industrial Engineering (UdL, 2015), an additional M.Sc. degree in Computer Vision (UAB, 2018), and a Ph.D. degree in Agricultural Technology (UdL, 2020). He currently serves as a researcher at the Institute of Agrifood Research and Technology (IRTA). His research focuses on the application of sensors, computer vision, and artificial intelligence in agriculture. He has leveraged this expertise to develop innovative methodologies for fruit counting and sizing. Furthermore, his recent work involves utilizing remote sensing data and artificial intelligence for crop mapping and agricultural water management.