コンテンツにスキップ

How Handheld LiDAR Is Changing Forest and Urban Tree Surveys

Walk through a forest once, then return to it on screen. Handheld LiDAR allows surveyors to capture the structure of a site as they move, creating a 3D record from which individual trees can be identified, measured and mapped.

What is a handheld LiDAR tree survey? A handheld LiDAR tree survey uses a portable mobile laser scanner to capture three-dimensional data while the operator walks through a forest, park or streetscape. The resulting point cloud can be processed into individual tree records containing location and structural attributes such as diameter at breast height (DBH), height and crown dimensions.

In brief:

  • Handheld LiDAR records the geometry around the operator rather than only one measurement at a time.
  • The point cloud can be revisited after fieldwork and processed into individual tree records, maps and reports.
  • Automated segmentation and measurement can reduce repetitive processing, but the results still need professional review.
  • The method supports both forestry and urban tree work, although the survey design and final decisions are different.

From individual readings to a record of the site

Conventional tools remain important in tree surveying. A diameter tape, calliper, rangefinder or total station can provide a clear measurement when the target is accessible and the survey design is well defined. The limitation is not necessarily the quality of an individual reading, but the amount of site context that can be recorded during the same visit.

A handheld LiDAR scanner collects spatial data continuously as the operator moves. Tree stems, ground surfaces, crown structure and nearby objects become part of the same point cloud. Instead of leaving the site with a list of isolated observations, the surveyor can leave with a three-dimensional record that can be inspected from different viewpoints.

This does not make the survey automatic. It changes where the work happens. More information is captured in the field, while tree identification, quality control, measurement and reporting continue in the software.

For an independent arborist or small consultancy, this can reduce the time spent reconciling separate measurements, notes and photographs after site work. Environmental consultants and land-development teams can review the same spatial record before a report, design decision or planning submission moves forward. Municipal tree-management teams can link individual tree records to locations and future maintenance information. For dealers and service providers, the value is a clearer workflow from field capture to a reviewable inventory and customer handover.

What does handheld LiDAR change in practice?

The most important change is not simply speed. It is the ability to connect a measurement to the tree and the tree to its surroundings.

Survey taskWith individual field measurementsWith a handheld LiDAR dataset
Record the siteMeasurements and notes describe selected trees or plots.The point cloud preserves the captured geometry and spatial context.
Review a treeA missing observation may require another site visit.The team can revisit the captured data when the relevant surface is visible.
Map the inventoryTree positions may come from a separate positioning step.Tree records can be connected to a georeferenced point cloud when the project uses an appropriate positioning or control workflow.
Check the resultQuality control focuses on recorded values and field notes.Measurements can be checked against the tree geometry, capture coverage and segmentation.
Create new outputsOnly the attributes collected in the field are available.The same dataset may support additional measurements, maps or visualisations, provided the required geometry was captured.

Handheld LiDAR forestry survey workflow: capture a tree plot, process the point cloud, extract tree insights and deliver a forest survey report


A point cloud is not yet a tree inventory

Capturing the site is only the first part of the job. A raw point cloud does not identify which points belong to the ground, where one tree ends and another begins, or whether a slim vertical object is a stem, a support pole or something else.

To become an inventory, the data needs to be organised into individual tree records. A typical workflow includes defining the analysis area, extracting the ground, segmenting trees, checking the result, calculating selected properties and preparing the final map or report.

What can be measured, and what must be calculated?

It is important to distinguish structural attributes derived from the captured geometry from estimates that depend on external models and assumptions.

Inventory fieldHow it is usedImportant condition
Tree ID and positionMaps each tree, connects records to the site and supports future identification.Project coordinates require an appropriate GNSS or control workflow.
DBHSupports forest mensuration, size-class analysis and downstream volume or biomass calculations.The trunk must be sufficiently visible at the project-defined breast height.
Tree heightSupports stand structure, growth and volume analysis.Dense canopy, hidden treetops and incomplete ground data can affect the result.
Crown dimensionsSupports spacing, canopy structure and growing-space analysis.Crown separation and point-cloud coverage must be reviewed.
SpeciesSupports classification and species-dependent calculations.Species is not determined by geometry alone. Imagery, field evidence and expert confirmation may be required.
Timber volume and carbon storageSupports resource estimates and selected reporting workflows.These are calculated estimates. Results depend on species information, selected models, parameters and supported forest conditions.

The table separates captured geometric attributes from species-dependent and model-based estimates. It also states visibility and control conditions for position, DBH, height and crown dimensions.

What happens after the surveyor walks the site?

ForestMind provides one example of how a handheld LiDAR dataset can move from capture to a usable inventory. The process in FJD Trion Scan and FJD Trion Model follows eight main stages:

Capture the survey area

Walk the planned route and collect the point cloud in FJD Trion Scan. The route, device configuration and positioning method need to match the site and intended deliverable.

Clip the point cloud to the analysis area

In the FJD Trion Model, use the clipping tools to retain the plot, tree group or corridor needed for analysis. Removing unrelated data makes the next processing steps easier to review.

Extract the ground

Choose the ground type that best matches the site, such as flat ground, a gentle slope or a steep slope. Ground extraction creates the vertical reference needed for tree segmentation and height-related calculations.

Segment the point cloud into individual trees

Select the appropriate woodland type and run tree segmentation. ForestMind assigns individual IDs and separates the point cloud into tree records that can be inspected and edited.

Review and correct the segmentation

No automatic result should be accepted without checking it. Users can delete objects incorrectly classified as trees, add missed trees and merge tree sections that belong to the same stem. In managed urban sites, support poles can also be removed before property calculation.

Calculate crown and tree properties

After reviewing the tree structure, run crown analysis and select the required properties. ForestMind can calculate values such as DBH, tree height and crown dimensions from the processed tree point clouds.

Tree species AI Identification

Species can be entered manually by comparing the point cloud with panoramic imagery. Where the AI suggestion function is available, the software can propose a species and display a confidence level. The suggestion remains a review aid.

Generate the final deliverables

Once the inventory has been checked, FJD Trion Model can export distribution maps, tree-level parameters, summary statistics, charts, and other project outputs for reporting and handover.



Where handheld LiDAR can change forest inventory

In plot-level forest inventory, handheld LiDAR can create a spatial record of stems, ground and visible crown structure within the surveyed area. After segmentation, the dataset can be organised by tree ID and used to examine tree counts, DBH classes, height distribution and spacing.

For a public-sector overview of how terrestrial and airborne LiDAR can support forestry inventory and tree metrics, see the USGS Interagency LiDAR Monitoring & Research Applications overview.

The value becomes clearer when the team needs more than a table. A tree can be viewed in relation to neighbouring stems and terrain, while distributions can be checked against the underlying point cloud. The captured geometry may also support later questions that were not part of the original field sheet, provided the relevant surfaces are visible in the data.

Volume, biomass and carbon values should be treated differently from DBH or height. They are model-based estimates and depend on the selected equations, species information and survey conditions. In the current ForestMind workflow, volume and carbon functions are primarily intended for single-species stands.

Why the same method matters in cities

An urban tree rarely exists in isolation. It stands beside roads, paths, buildings, utilities, street furniture and other managed assets. Mobile laser scanning can capture part of that context during the same survey, making the resulting inventory more useful for mapping and future site review.

Each tree record can be linked to a position, selected structural attributes and, where captured, panoramic imagery. This can support parks, campuses, roadside inventories and public green-space management. It does not, however, turn geometric data into a condition or risk assessment. Those conclusions still belong to a qualified arborist working to the required local standard.

For professional tree-risk assessment context, refer to the ISA Tree Risk Assessment Qualification.

How does a tree survey become GIS-ready?

A tree inventory becomes GIS-ready when each record has a reliable position, a defined coordinate reference system and a consistent set of attributes. A handheld LiDAR survey can support this output when capture is connected to an appropriate positioning or control workflow and the processed tree records can be exported in the required format.

GIS-ready tree data can help teams view tree distribution by area, connect inventory records with other land or asset layers, and plan future site work. The point cloud and tree table should still be checked before they become part of an authoritative asset database. 

Where browser-based review is useful before handover, FJD Trion Model Web can help project participants view, measure and share point-cloud data. Confirm current access settings, supported formats and data-management requirements before using it as the project review environment.


How reliable is LiDAR-derived DBH?

LiDAR can support DBH measurement when the trunk is captured with enough detail and the correct section of the stem is identified. Accuracy is influenced by point density, bark shape, vegetation, occlusion, positioning and the fitting method used by the software.

An independent 2026 ISPRS study evaluated handheld mobile laser scanning for DBH estimation in an urban park using two systems, including the FJD Trion S1. The study reported DBH RMSE values of 3.3–3.6 cm for filtered datasets and found that point density affected tree detection. These results belong to that specific test and should not be treated as a universal accuracy specification. They show why capture quality and result checking matter. Read the study.

Automation changes the workload, not the responsibility

Tree segmentation and property calculation can reduce repetitive manual work, but neither should be treated as an unquestionable result. Occlusion, dense understory, touching crowns, split stems and non-tree objects can all affect processing.

Software can supportA professional still needs to
Ground extraction and individual-tree segmentationCheck missed trees, false detections and incorrectly merged or split stems
DBH, height and crown property calculationConfirm the correct measurement convention, data quality and acceptable uncertainty
Manual and AI-assisted species labellingVerify species when the result affects management or calculation
Distribution maps and structured reportsApprove the survey scope, professional conclusions and final handover
Volume and carbon calculations in supported conditionsSelect and validate the correct species parameters, models and reporting method

Where handheld scanning reaches its limits

A handheld scanner can only measure surfaces reached by the laser. Dense vegetation may hide stems. Overlapping crowns may be difficult to separate. The top of a tall tree may not be captured clearly from the ground. Poor route planning can leave gaps, while weak positioning or control can limit the reliability of mapped tree locations.

The method is therefore strongest when capture planning and processing are treated as part of the survey method. Device range, point density, terrain, vegetation, access, positioning and the required output all influence the result. Larger or less accessible areas may require backpack, vehicle, aerial or combined capture rather than a handheld-only approach.

The real shift: one field visit, a reusable spatial record

Handheld LiDAR does not replace forestry knowledge or arboricultural judgement. Its contribution is different. It allows the surveyor to bring back a spatial record of the site rather than only the answers written down in the field.

That record can be segmented into trees, checked against the visible geometry, mapped and turned into structured inventory data. The survey remains a professional process, but the evidence available to the professional becomes richer, more visual and easier to revisit.

Explore FJD Trion Tree & Vegetation Management and FJD Trion customer case studies for related workflows and published examples.  For your next tree-inventory project, talk to the FJD Trion team about the right capture method, review process, and handover requirements.

Frequently asked questions

Learn how FJD Trion combines handheld LiDAR capture, point-cloud processing, and the ForestMind tree-inventory workflow to support forest and urban tree surveys, structured inventory records, GIS-ready data, and site documentation. Explore what LiDAR can measure, what still requires professional judgement, and what teams should review before final handover.

No. LiDAR does not replace professional judgement, but it can reduce repetitive manual measurements and field-record consolidation. A qualified professional still defines the survey method, checks the data and makes decisions about species, condition, risk, management and regulatory compliance.  

No. Automatic segmentation can produce false, missed, split or merged detections, particularly in complex vegetation. The operator needs tools to delete false detections, add missed trees and merge sections that belong to one tree before calculating properties.

LiDAR primarily records geometry. Species identification normally requires additional evidence such as imagery, field observations or expert knowledge. Software may suggest a species, but the result should be confirmed when it affects management, volume or carbon calculations.

Yes. The core capture and processing workflow can support forest plots, parks, campuses, roadsides and other managed tree sites. The required attributes, positioning method, professional standard and final deliverable will differ by project.

Depending on the survey scope, it can contain tree IDs, positions, DBH, height, crown dimensions, species records, summary statistics, distribution maps and links to imagery or other asset data.  

No. LiDAR records geometry. Carbon storage is estimated using measured or derived tree attributes together with species information, equations and selected reporting methods. The assumptions and supported conditions need to be stated with the result.  




アーカイブ