At a UK, European and Global scale there is clear public desire and political ambition to reverse environmental degradation and create a 'nature-positive' future, as shown by the UK Government's 25 Year Environment Plan and the Convention on Biological Diversity Montreal Framework. In the UK, key mechanisms to unlock green investment on a scale necessary to deliver on this ambition include the emergence of carbon financing linked to landscape restoration and mandatory Biodiversity Net-Gain for almost every scale of development. However, in order to unlock the potential of these initiatives, there remains the need for more effective landscape-scale evaluation of habitat condition. When describing habitats and ecosystems we tend to think of them merely as assemblages of species, but generally they are more complex than that because important micro- or macro-scale features also play key roles in the character and functioning of the habitat. There is growing recognition that habitat structure is often just as important as species descriptions, yet the tools required to describe such structures and features are poorly developed compared with long-established methods of identifying and describing species and species assemblages. Peatlands and 'open mosaic habitat' both in their respective ways rely on habitat structure for their character and ecological functions as much as they do on their species complement. Such structures give rise to textures or patterns that are often most easily seen from above, but consistent and systematic descriptions of these textures is less easily achieved because the vocabulary for pattern description is poorly developed. AI machine-learning of high-resolution remotely sensed imagery, however, offers a means to do this by identifying texture-pattern types and then classifying them according to habitat type and condition. For this to be effective, however, there needs to be active and close engagement between ecology and AI-led image analysis. This project seeks, through such engagement, to generate a new approach to habitat identification and condition assessment based on the texture-patterns generated by the combination of species assemblages with their associated structures, thereby creating a system of habitat 'fingerprints' that can be used to map habitat character and condition. For peatlands, this offers the opportunity to map the condition of entire landscapes and thus underpin market initiatives such as the Peatland Code, while also enabling, at the small scale, the mapping and characterisation of open mosaic habitat for obligatory actions such as Biodiversity Net Gain.