LiMapper Lite 1.1.2 | User Guide

Introduction

LiMapper is an aerial image processing software for mapping and photogrammetry. It provides the ability to automatically stitch images taken from airborne devices including fixed and rotary winged UAV.

By utilizing computer vision techniques, LiMapper identifies matching points across overlapping images. When using images from varying angles, LiMapper uses photogrammerty methods to determine the relative position of the camera sensor, allowing for the generation of 3D surface models such as DOM (digital orthophoto map) and DSM (Digital surface model).

Easy to use and performance driven

LiMapper has solved traditional mapping issues, including altitude instability and distortion from aerial imagery. It supports traditional and oblique imagery and has the capability to process 10’s of thousands (>10,000) of images simultaneously. LiMapper provides a very friendly user interface for automated workflows, suitable for advanced users and beginners alike.

Suggest edit
Last updated on August 22, 2016

Installation

Operating Environment

RAM: 16GB+;

CPU: Intel(R) Core(TM) i5/i7; dual-core processor;support SSE3.

GPU: NVIDIA CUDA supported GPU with at least 2GB RAM; support GLSL.

Operating System: Microsoft Windows 7 (64 bit), Microsoft Windows 8/8.1 (64 bit), Microsoft Windows 10 (64 bit).

Note: Adjust the graphics mode to high performance.

Installation

Download the LiMapper installer from the Greenvalley website and run it (LiMapper Lite Setup.exe).

License Manager

The LiMapper installer will install a trial version of the software that is active for a limited time with limited functionality (100 images may be loaded using the trial version). To activate LiMapper, first purchase a license by contacting us at info@greenvalleyintl.com.

After purchasing a license, activate it by following these instructions.

1. Run LiMapper as administrator by navigating to Start > All Programs > LiMapper Lite (Right click LiMapper Lite > Run as Administrator).

2. Go to license activation by navigating to Help > License Activate (if your trial has expired, this window will automatically appear). 

image20

3. Fill in the required fields – Name and Company, denoted by the asterisk (*). Clicking Copy for E-mail will copy in information needed for activation to the clipboard. 

image13

4. Compose a new email to support@greenvalleyintl.com with the subject “LiMapper License Activate” and paste the information copied earlier into the body.

5. A License Key will be generated for you within 24 hours. Once it is available, finish the activation by filling in the License Key field with the license key.

Note: If any issues are encountered during the activation process contact support@greenvalleyintl.com.

Suggest edit
Last updated on August 22, 2016

Getting Started

LiMapper is designed to be straightforward and easy to use. Below is a brief overview and summary of the application.

Suggest edit
Last updated on August 22, 2016

Application Overview

Main application window.

image01

Title Bar: Displays the application name (far left) and Windows controls (far right)

Menubar: All tools, settings and application controls.

Toolbar: Shortcuts to common tools and settings

Workstation Panel: List of all loaded files and models

Ground Control Points (GCP) Panel: List of ground control points

Photos Panel: List of loaded photos

Console Panel: Executed command list

Suggest edit
Last updated on August 22, 2016

File Menu

File operations, including loading, saving and creating projects.

New Project

File > New…

Create a new project. Required to give a project name and a save location.

Open Project

File > Open…

Open an existing project.

Save

File > Save

Save changes of the loaded project.

Save as

File > Save As…

Save changes of the loaded project to a new file (or overwrite an existing one).

Generate Quality Report

File > Generate Quality Report

Generate a detailed report in the PDF format.

Export 3D Points

File > Export 3D Points

Export the point cloud generated from the photogrammetry model.

Export 3D Mesh

File > Export 3D Mesh

Export the mesh generated from the photogrammetry model.

Open Product Path

File > Open Product Path

Opens the directory containing the ortho mosaic product.

Exit

File > Exit

Exits the application.

Suggest edit
Last updated on August 22, 2016

Workflow Menu

The main procedure for executing photogrammetry processing.

Add Photos

Workflow > Add Photos…

Add the photos to participate in the photogrammetry processing.

Blend Photos Procedure

Workflow > Blend Photos Procedure…

Photogrammetric processing; blends/stitches photos together by automatically identifying control points and calculating the camera projections.

Suggest edit
Last updated on August 22, 2016

Tools Menu

Import

A collection of import tools (currently, only control points can be imported).

Control Points

Tools > Import > Control Points…

Import pre-existing control points.

Export

A collection of export tools.

Control points

Tools > Export > Control Points…

Export LiMapper’s calculated control points.

Camera Orientations

Tools > Export > Camera Orientations…

Export LiMapper’s calculated camera orientations.

Undistorted Photos

Tools > Export > Undistorted Photos…

Export the undistorted version of the individual photos (may take some time depending on the number of photos loaded).

Undistort

Tools > Undistort…

Undistort photos using manual parameters in a batch-like process.

Bounding Box

Manually change the bounding box in the visualizer.

Resize Region

Tools > Bounding Box > Resize Region

Resize the bounding box in the visualizer using the corner points.

Rotate Region

Tools > Bounding Box > Rotate Region

Rotate the bounding box in the visualizer.

Settings

Tools > Settings…

Change global settings for LiMapper, including the number of central processing units (CPUs; defaults to the number of cores available on the system) and the graphics card (GPU) to use.

Suggest edit
Last updated on August 22, 2016

View Menu

Controls the look and feel of LiMapper.

Show Feature Points

View > Show Feature Points

Show/hide feature points.

Show Trackball

View > Show Trackball

Show/hide the trackball in the visualizer.

Show Bounding Box

View > Show Bounding Box

Show/hide the bounding box in the visualizer.

Show GPS

View > Show Feature Points

Show/hide GPS data.

Show Camera

View > Show Camera

Show/hide camera locations and orientations.

Show Ground Control Points

View > Show Ground Control Points

Show/hide points used for ground control points.

Increase Camera Size

View > Increase Camera Size

Increase the projection visualization of the cameras.

Decrease Camera Size

View > Decrease Camera Size

Decrease the projection visualization of the cameras.

Show Point Cloud, Tri-Mesh, Model

View > Show [Point Cloud, Tri-Mesh, Model]

Choose how to display the calculated model.

Show Sparse/Dense Cloud

View > Show [Sparse, Dense] Cloud

When displaying as a point cloud, choose whether to show as a sparse or dense point cloud (for the latter, it must be calculated as such).

Ortho Render

Display the model with an ortho projection.

Suggest edit
Last updated on August 22, 2016

Windows Menu

Controls how the windows within the application are displayed.

image12

Toolbar

Show/hide toolbar.

Workstation

Show/hide workstation panel.

Ground Control

Show/hide ground control panel.

Photos

Show/hide photos panel.

Console

Show/hide console.

Suggest edit
Last updated on August 22, 2016

Help Menu

image04

View Help

View this user guide.

License Activate

Activate the license for LiMapper.

About LiMapper

Information about LiMapper including the version.

Suggest edit
Last updated on August 22, 2016

Workflow

LiMapper makes image stitching and surface model building really simple. This consists of 2 essential steps: adding photos and the blend photos procedure (both found in the workflow menu). The blend photos procedure is the core of the data processing; please refer to the Optimizing the Blend Photos Procedure chapter for more information about its parameters.

Here, interface elements, common workflow and data visualization will be outlined.

Suggest edit
Last updated on August 22, 2016

Data Management

Import and view data.

Beginning a new project and adding photos

Before adding data, a new project must be created (or an existing project must be opened). See the “New project” in the “Getting started” section for more information about projects. Navigate to File > New Project to create an empty project. This will open the new project dialog (see figure below).

image08

For the project name, it is customary to name it after the location the images was taken (however, it can be anything; this is for a user’s own referencing purposes). For project location, use the ellipsis (…) button to set a path to the project (this will be where a project is saved, including any data produced during data processing).

Add Photos (and Remove)

After a project has been initialized, navigate to Workflow > Add Photos. This will open a dialog allowing for image selection; select all the photos that will participate in the procedure and click “Open”.

It may take a few moments to load the images (depending on how many photos were selected); the status can be viewed in the status bar. When completed, a list of loaded images can be viewed from the workstation window.

image06

Remove Photos

Photos can be removed from a project; From the workstation window, right click the photo to be removed and select “Remove Photo”.

Photo Visualization

Each photo can be viewed individually by double clicking on the photo from the workstation window. The image will open in an adjacent tab to the model windows in the viewer window.

image00

(Additionally, the image will also be visible at the bottom of the workstation window). The close button can be used to close the viewing window (note: this only closes the viewer; it does not remove the data).

Zooming

Zoom in and out of a photo by clicking in the viewer and using the mouse scroll. Additional, a specific extent can be zoomed into by clicking and dragging the extent over the image.

To reset the image zoom (fit to viewer), click the home button on the toolbar.

image23

Panning

To pan an image (move left, right, up and down), use the mouse middle click and drag.

Suggest edit
Last updated on August 22, 2016

Blend Photo Procedure

To start image stitching and photogrammetry analysis, navigate to Workflow > Blend Photos Procedure. This will be a brief overview of the procedure, refer to the Optimizing the Blend Photos Procedure chapter for explanations of parameters and performance optimizating.

image11

In the dialog, the mode of image collections can be selected:

  • Ortho Mode – Choosing this will attempt to project all photos to a 3D model for blending. This is suitable for UAV aerial photography with a random flight.
  • Planer Mode – Choosing this will attempt to project all photos to a ground plane for blending. This is suitable for traditional large-format aerial photography with regular flight.

Many times, either option will succeed for a dataset. However, there may be an impact on the results’ performance when using a less preferred method.

Blending

Refer to the “Optimizing the Blend Photos Procedure” chapter for a detailed determination of the blending mode and for parameter explanation and optimization.

In most cases, accepting the default parameters and clicking OK, Next and Finish (where appropriate) will start the image blending process (for either mode). If successful, the result will appear in the Model and Product subsections of the workstation window and be displayed in a model viewer as a sparse set of surface points (these are the tie points used to tie the individual images) and calculated camera positions, similar to the following figures.

image15

Figure: Model results listed in the workstation window

image22

Figure: Model displayed as tie points and relative camera positions.

Visualizing the 3D Model

A model can be viewed from the model viewing window. At any time, this viewer can be reset to the home position by clicking the “Home” button on the toolbar.

image23

Rotating

Click the left mouse button and drag to rotate the model for viewing at different angles.

Zooming

Use the mouse scroll to zoom in and out of a model.

Panning

Click the middle mouse button and drag to pan the model.

Individual Camera Projection

Right clicking the model viewer will cause individual camera project to become active. This projection allows the viewing extent to become visible for easily identifying the zone for each image.

Display Mode

There are 3 display modes for a model: Tie Points, Tri-Mesh and Surface Model. Note: Models derived from planer mode blending only results in tie point viewing.

Tie Points: View > Show Point Cloud (additionally, you can double click the Tie Points model in the model subsection of the workstation window). This mode displays the tie points used for blending the individual photos.

image25

Tri-Mesh: View > Show Tri-Mesh (additionally, you can double click the Tri-Mesh model in the model subsection of the workstation window). This mode displays the triangular mesh surface derived from the ortho mode blending.

image18

Surface Model: View > Show Model (additionally, you can double click on Model in the model subsection of the workstation window). This mode displays the surface model derived from the ortho mode blending.

image26

To change between the different display modes, navigate to the View menu and select either

Miscellaneous Options

Camera Size: Increase and decrease the size of the camera projection within the model viewing window. Navigate to View > Increase Camera Size (or Decrease Camera Size). Alternatively, these controls can be found on the toolbar.

image07

Suggest edit
Last updated on August 22, 2016

Optimizing the Blend Photo Procedure

The blend photo procedure offers 2 methods of blending, depending on the mode of image collections.

Suggest edit
Last updated on August 22, 2016

Determining the Image Collection Type

Some information is needed about the how the images were collected. There are generally just 2 modes: Ortho and Planer.

It is relatively simple to choose the correct collection mode; consider the following 2 cases.

Case 1 (Ortho): If photos were taken from a camera attached to something like a UAV, which frequently changes its yaw, pitch and roll or has an array of cameras, such that overlapping photos are taken at different angles, ortho should be chosen.

Case 2 (Planer): If photos were taken from a camera attached to something like a fixed winged, large aircraft, which follows a straight planer path, such that overlapping photos are all taken perpendicular to the surface, planer should be chosen.

To begin the blending process, navigate to Workflow > Blend Photos Procedure.

Note: this assumes a project has been set up and photos have already been added. If this hasn’t been done yet, refer to the Workflow chapter to get started.

Suggest edit
Last updated on August 22, 2016

Procedure Outline

Above the mode selection is the procedure outline. This is slightly different for each mode, but they work in the same way. The outline displays the steps performed during the blending process.

image02 image17

Figure: Outlined steps for Ortho and Planer Mode, respectively.

These steps are performed in order and each step relies on the previous step’s results. Each step has its own set of parameters (that may be set in the following dialogs). If it is the first time running a blend, it is required to perform all outlined steps before producing the final model. However, each step can be performed one at a time (i.e., execute only “Align Photos”); these results will be saved and can be used at anytime by the succeeding step. The following figure shows latter step selection for a blend procedure when a former procedure has already been executed.

image11

This allows for changing of any particular step’s parameters without having to redundantly carry out any previous steps; it would just reuse any previous steps results.

Note: when changing the parameters of any step, all succeeding steps must be re-executed before producing a new blending result.

Note: the align step is the same for both modes, and the results can be used interchangeable.

Suggest edit
Last updated on August 22, 2016

Ortho

In ortho mode, there are 3 steps for blending: align photos, build model and finally bending the photos.

image02

Pressing OK will cause a series of dialog windows to appear. The following sections detail these parameter dialogs.

Align

Set alignment parameters in the “Align Photo Settings” dialog window. This alignment produces the tie points when building the model.

image05

Detection

Detection sensitivity can be set using High, Medium and Low and the maximum number of features can be set. Higher values may yield better accuracy, however it will increase the alignment time and a higher use of resources.

Matching

Pair selection mode can be set to generic or exhaustive and the maximum tied points can be set. Higher values may yield better accuracy, however it will increase the alignment time and a higher use of resources.

Bundle Adjustment

Different fitting analysis can be excluded if necessary (all are used by default).

Build

The model building parameters will be set in the “Build Model Settings” dialog window. The model building builds the Tri-Mesh and 3D surface model, models.

image21

Source

There are 3 options for creating the source point cloud: Sparse, Dense (high) and Dense (low). Selecting sparse will result in a quickly generated point cloud, whereas selecting dense will result in a denser point cloud (to potentially increase accuracy). Choosing Dense (high) will generate the most dense point cloud from the 3 options.

Quality

There are 3 options for the quality of the point cloud: Low, Medium and High. Higher values may yield better accuracy, however it will increase the generation time and a require a higher use of resources.

Blend

The blending parameters will be set in the “Blend Photos Settings” dialog window. This step creates the final mosaic.

image19

Projection Plane

Choose a method for projection used to start generating the model.

The whole model on its estimating plane – use an estimated plane to generate the model (default).

The whole model on the [X-Y, X-Z, Y-Z] plane – use one of the 3 simple projections planes/faces to generate the model.

The whole model on the current view plane – use the plane for the current view in the visualizer to generate the model.

The selected model in the bounding box on show – use the plane of the selected model to generate the model.

Sample Distance

The distance used to generate the model.

Output DSM

Choose whether or not to output the DSM during the blending procedure.

Suggest edit
Last updated on August 22, 2016

Planer

In planer mode, there are 2 steps for blending: align photos and bending the photos.

image17

Pressing OK will cause a series of dialog windows to appear. The following sections detail these parameter dialogs.

Align

Set the alignment parameters in the “Align Photo Settings” dialog window. This is exactly the same as align in ortho mode (and as such, can be interchanged between the 2 modes). Please refer to the Align section of the ortho mode for an explanation of settings.

Blend

The blend settings can be set in the “Blend Photos Settings” dialog window.

image03

Projection Plane

Choose a method for projection used to start generating the model.

The whole model on its estimating plane – use an estimated plane to generate the model (default).

The whole model on the [X-Y, X-Z, Y-Z] plane – use one of the 3 simple projections planes/faces to generate the model.

The whole model on the current view plane – use the plane for the current view in the visualizer to generate the model.

The selected model in the bounding box on show – use the plane of the selected model to generate the model.

Distance and Dimension

Simple Distance – the distance used to generate the model.

Max Dimension – maximum dimension to use to generate the model.

Filter Oblique Images

Filter out images with values larger than the angle indicated by this parameter.

Suggest edit
Last updated on August 22, 2016

Exporting

There are a few products and methods for obtaining the results processed after completing a workflow.

Suggest edit
Last updated on August 22, 2016

Mosaic Product

This is the assembled product created during processing. It can be found in the workstation window under “Product” and viewed by double clicking it.

This product is located in the product folder in the project’s directory, but can easily accessed within LiMapper by right clicking the product from the workstation window and choosing “Open Product Path” (see figure below).

image10

Suggest edit
Last updated on August 22, 2016

Export as a Point Cloud

To export the generated product as a point cloud, navigate to File > Export Points. This will open a dialogue allowing additional attributes to be saved in the point cloud (i.e., the color value and point normals for individual points).

image14

Note: The type of cloud available to “Source data” depends on parameters used to generate the cloud during the processing workflow. Refer to the build step in ortho blending of the Optimizing the Blend Photos Procedure chapter.

Point clouds can be saved in the PLY and OBJ formats to be imported in other programs such as MeshLab.

Suggest edit
Last updated on August 22, 2016

Export as a Mesh Model

To export the generated product as a mesh model, navigate to File > Export Mesh. Point clouds can be saved in the PLY and OBJ formats and can be opened in other programs such as MeshLab.

Suggest edit
Last updated on August 22, 2016

Appendix

Key Terms

Photogrammetry – Use of photographs in mapping and survey to determine distances.

Ortho – the projection of a 3D model onto a 2D surface (i.e., displaying a point cloud on a computer monitor such that points further away look closer together).

Tie Points – Similar points shared process different photos that are used to align photos for stitching.

Suggest edit
Last updated on August 22, 2016
Suggest Edit