Everyone wants insight into trees. Municipalities want to monitor for greening, grid operators want to minimise risks and policymakers are looking for rationale for climate measures.
But insight is only valuable if the data is reliable.
NEO delivers top-quality tree data and we are happy to explain why. We develop in-house AI technology that detects, analyses and tracks trees with unprecedented precision.
And that on a national scale.
Proprietary AI and deep learning: not a black box, but purpose-built
Our tree data is the result of self-developed deep learning models trained on tens of thousands of examples in diverse environments: from urban neighbourhoods to forested areas.
This allows our algorithms to understand what is a tree and what is not.
Indeed, our models recognise:
- Tree height
- Stemposition
- Tree type
- Undergrowth
- Crown volume and diameter
- Vulnerable tree zone
- Root risk zone
Powerful combination of data sources
Our strength is not only in smart models, but also in the breadth of what we put into them:
- Satellite imagery: continuous and large-scale overview
- Aerial photos: high resolution, multiple angles of view
- AHN (Actueel Hoogtebestand Nederland): millimetre precision in height and structure
By combining these sources, we build a 3D understanding of each tree. Not just what you see, but what you wouldn't otherwise can see.
But we go beyond that...
Data quality through smart filtering and context understanding
A major reason why our tree data is so reliable lies in the way we deal with disturbing elements. The Dutch landscape contains all kinds of objects that make it difficult to detect trees. Think of high-voltage power lines or covering buildings, for example.
Many standard models confuse these structures with tree tops or crowns, leading to noise or false detections. Or they cause nothing to be detected at all.
NEO actively prevents this by explicitly identifying and filtering out these objects in our analyses.
This contextual understanding increases accuracy and ensures that our data remains reliable even in complex conditions, such as in densely built-up neighbourhoods or under infrastructure.
Scalable and flexible
Whether you want to analyse a city district or the whole country: the technology scales with it. Because everything is AI-driven, we can quickly deliver up-to-date datasets at the neighbourhood level, municipal level or for the whole country.
Applications with impact
Our tree data service has two flavours: TreeBase and TreeRisk.
TreeBase provides detailed and up-to-date information on all trees in the Netherlands. This data forms a foundation for organisations and governments that want to substantiate policy, measure performance or target greening. Think of applications such as:
- Monitoring greening and climate adaptation
- Substantiation of policy objectives or grant applications
- Understanding crown volume and shading for livability analyses
- Public space management and maintenance support
TreeRisk, developed based on the same data quality, goes one step further: it maps tree-related risks. By cleverly linking trees to objects such as buildings, roads and cables, Tree Risk makes risks predictable and manageable.
Examples include:
- Storm damage risk analysis for insurers and governments
- Understanding root risks along roads, cabling or pipes
- Prioritisation of management measures based on potential damage or nuisance
Together, Tree Base and Tree Risk offer a complete spectrum: from factual inventory to risk-based action.
Quality does not come naturally
NEO's tree data is not just 'a map of trees'. It is the result of years of innovation, in-depth modelling and clever combinations of reliable sources.
Want to make a difference with the most reliable tree information? Take a look at TreeBase and TreeRisk.