Syllabus

• Week 1

  Introduction to image analysis for plant phenotyping

  • Welcome and Introductions

    Welcome to the course.

  • What is Image analysis?

    What do we mean by image analysis? And how do we apply it to typical plant phenotyping problems?

  • Typical image analysis problems in plant phenotyping

    In plant phenotyping you often want to count or measure some feature of a plants physical shape. An overview of the ways in which image analysis can help you do this, including image segmentation, object and feature detection.

  • Images as data

    We will discuss exactly what digital image data consists of, and what you need to consider to get high quality digital images for use in your analysis.

  • Image Formats and Data Types

    A quick run through of common digital image formats, with advice on what format to use in different contexts

  • Summary and review

    What have we learned so far? And what will we look at next week?

• Week 2

  Tools for image analysis

  • Introduction to Fiji (ImageJ)

    We focus on one tool in particular, Fiji (also known as ImageJ). We show how do perform simple tasks such as loading images, measuring image features and batch processing. We also introduce ImageJ scripting.

  • Image thresholding

    A common task in image analysis is binary thresholding, or dividing an image into two regions representing some foreground object and a background object. In this activity we look into various ways we can do this.

  • Python basics

    This activity will introduce you to the basics of running Python code on your computer, and show several different ways to do so. We will also show the basics of navigating using a command line, and how to install Python packages.

  • Summary and review

    What have we learned so far? And what will we look at next week?

• Week 3

  Coding for image analysis

  • Introduction to coding for image analysis

    An introduction to programming using Python, with particular emphasis on its use in image analysis. We introduce functions, variables and data types, and show how Python can be used within Fiji.

  • The building blocks of programming

    An introduction to the basics of Python programming, including For and While loops, and decision making using If statements

  • More complex programs: Dilation and Erosion

    Now we have more programming building blocks in Python we can start piecing them together to perform simple image analysis tasks in Fiji. In particular we look at the morphological operations dilation and erosion.

  • Summary and review

    What have we learned so far? And what will we look at next week?

• Week 4

  Common tasks in image analysis

  • Noise reduction

    An introduction to the types of noise envountered in image data and how to reduce it. Covers convolution techniques such as Gaussian filtering.

  • Contrast enhancement

    A look at how image clarity can be improved via contrast enhancement, including the use of histogram equalisation.

  • Counting and labelling via Segmentation

    A look at how to identify and quantify image features. Covers image segmentation, including classification, clustering and spatial grouping approaches

  • Feature-based methods

    Often we seek to find point or edge features in our images rather than regions. We look at edge and corner detection using convolutional methods such as the Sobel filter.

  • Model-based approaches

    What do we mean by model-based approaches to image segmentation? We look in detail at the example of Active contours, also known as Snakes

  • Summary and review

    A quiz and summary of the week's material

• Week 5

  Beyond individual 2D images

  • Video data and motion detection

    A look at common image analysis techniques used when dealing with sequences of images or video, including pixel-based methods, motion detection and background subtraction

  • Measuring motion and object tracking

    We continue our look at motion in video data by first seeing how motion of pixels is measured using optic flow. Next, we look at how objects are tracked using model-based methods such as active contour models

  • Volumetric images

    An introduction to capturing, viewing and analysing volumetric image data

  • Other image data types

    Our look at image data types beyond simple 2D concludes by looking at mosaicing of aerial images, and hyperspectral and multispectral images

  • Summary and review

    A review and summary of 'Beyond 2D images'

• Week 6

  3D image reconstruction and course summary

  • 3D imaging in plant pheotyping

    What 3D imaging techniques are available and why are they useful in plant phenotyping. We discuss common data types and give examples of their uses.

  • Image based reconstruction

    An overview of the common methods of 3D image reconstruction

  • Case study: 3D reconstruction of plant canopies

    To round off our introduction to 3D image reconstruction we look at two related case studies from the University of Nottingham where 3D reconstruction techniques were used to model plant leaf canopies.

  • Discussion and next steps

    A discussion on what we have covered in the course, and how it feeds into more advanced topics such as machine learning.

Who will you learn with?