The ability to accurately characterise complex variation in organismal colouration has important implications for multiple fields of bioscience research, and there is a need to develop efficient and effective new workflows for harnessing the vast potential of image-based biological datasets. Recent advances in artificial intelligence (AI) and computer vision provide a powerful opportunity to achieve this, but software pipelines making these approaches generalisable and readily accessible to the bioscience research community have yet to be developed. In this project we will build on our track record of colouration, AI and computer vision research to develop a 'next generation' software toolkit for extracting and analysing high dimensional colour pattern information from images. These tools will be integrated into a user-friendly interactive software package that will have wide applicability across the biosciences and will transform the ability of researchers to rapidly characterise colour pattern phenotypes from image datasets. To achieve this the project is divided into three work packages split into developing advanced tools for segmenting and analysing complex organismal colour pattern variation from images and then implementing these tools in a user-friendly interactive framework. In the first work package, we will adopt a cutting-edge hierarchical semantic segmentation strategy to develop models that are capable of accurately detecting not only a specimen within an image but also of simultaneously segmenting regions within the specimen, thus enabling a detailed analysis of its constituent parts. In the second work package, we will develop a powerful new workflow for colour pattern analysis that leverages the potential of deep learning. This new pipeline consists of multiple steps, each employing cutting edge techniques, and will equip researchers with the power to efficiently perform advanced colour pattern analysis in their system. In the final work package, we will incorporate the new segmentation and analysis tools into 'Phenolearn', an easy-to-use Python-based software program developed by us for biological image analysis using AI. Together, these work packages will produce a powerful and generalisable toolkit for extracting and analysing colour pattern information from biological images. The availability of these tools via this project has the potential to catalyse existing bioscience research programmes and to open up new fundamental and applied research areas involving colour pattern phenotyping that are currently intractable.