Welcome to the course.

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

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.

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.

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

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

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.

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.

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.

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

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.

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

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.

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

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

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

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

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.

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

A quiz and summary of the week's material

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

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

An introduction to capturing, viewing and analysing volumetric image data

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

A review and summary of 'Beyond 2D images'

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.

An overview of the common methods of 3D image reconstruction

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.

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