Introduction to I-CONIC Studio

I-CONIC Studio is an easy-to-use tool to instantly create 3D point clouds from video or still imagery.

The purpose of the Free Trial is to get an immediate visual understanding of the regions-of-interest, to validate the recorded data, to make basic measurements and to easily export the point clouds to external software packages for advanced analysis and visualisation.

Licensing

During your trial you can enter your product key from the start view. Press the Register button at the sidebar and enter the product key.

When your trial has expired, a product key will be required to run the application. A dialog will show and ask for your product key when starting the application.

System requirements

Requirements
Operating system Windows 10 or 11
GPU Nvidia GeForce (10-series or better), updated driver routines, minimum 6 Gb

The Trial version works with the mentioned requirements, however, it may work with other GPUs. When you have opened I-CONIC Studio, you can get information about your GPU via image-20230619171401946View -> Info -> Help -> System Info*. We appreciate if you inform us about GPU issues and successes by sending a message to info@i-conic.eu.

In order to run I-CONIC Studio you need to accept the Licence Agreement. For the Trial version of Studio, the licence is embedded in the software you have received, and it will expire at the end of the trial period.

Requirements on video and imagery

All common video and image formats are accepted as input.

The foundation of photogrammetric processing is two pictures covering the same object from two different angles. In I-CONIC Studio, this means that the images selected from the image folder, or the video frames selected in the video file, have to cover the same area to some extent.

So, if a part of the video is, for example, from a drone hovering over the same area, then it is not possible to create a 3D model. Models begin to be created when the drone starts flying.

For the Trial version of I-CONIC Studio, problems with 3D modelling might occur if:

  • the images/frames in the pair are rotated in relation to each other (e.g. when the drone is turning).
  • there is very little contrast between, or inside, objects in the images/frames (e.g. areas with snow or water).
  • there are objects moving between the take of the first and second image/frame in the pair. Moving objects might disappear or disturb the model.
  • there is very large distance between the closest and farthest objects, e.g. when the images/frames are taken towards the horizon

Running video and image projects

When you open I-CONIC Studio, you will see this window.

image-20230705152328383

In the Drag-and-drop zone you will find some video sequences you may use for initial testing of I-CONIC Studio.

Create new project

There are several ways to create a new project after opening I-CONIC Studio, where #1 is the quickest option:

  1. Drag and drop a video file or an image folder, to the Drag-and-drop zone. The first video frames or images are ingested and the processing is ready to run.
  2. Click the image-20230616095053899 Video or image-20230616095123174 Image button to browse for your data, depending on what type you have. Click OK, and the first video frames or images are ingested and the processing is ready to run. If you have images, they will be processed sequentially in numerical order based on the image file name.
  3. If you are new to I-CONIC Studio, it could be beneficial to start with some test data. Download the data from examples in the drag-and-drop zone. Then open the test data with any of options #1 or #2 above.

Now, click the Play button or go to Settings for specifying processing parameters.

Open existing project

There are several ways to open an existing project after opening I-CONIC Studio:

  1. Drag an I-CONIC project file (*.icproj) to the drag-and-drop zone.
  2. Click the image-20230616101716065 Open button.
  3. Click any of the project links in the Recent projects list in the Drag-and-drop zone.

The (*.icproj) project files are readable text XML documents. You can, at your own risk, edit the parameters in the XML document. You can also add your own parameters to the file. I-CONIC Studio will read, write and rewrite all parameters within the file, but only use those that are defined by I-CONIC. Do not delete any I-CONIC parameters from a file.

Georeferencing

Georeferencing is the process that relates an internal coordinate system to a ground system with geographic coordinates. In order to georeference data there has to be metadata available.

I-CONIC Studio will generate 3D point clouds in an internal coordinate system if there are no or insufficient metadata. You can rescale the internal coordinate system to e.g. a metric scale by using Set scale under Measure.

If there are metadata, then the 3D point clouds will be georeferenced, saved in a geographic coordinate system, and visualised in the map window. If it is not possible to determine terrain height from metadata, then an external Digital Elevation Model (DEM) is used. You can download a low resolution global DEM under Settings/Options/Camera.

Georeferenced point clouds can be directly exported to a GIS and other software for analysis and fused with other georeferenced data layers.

The following metadata types are currently handled by Studio:

  • Imagery: DJI and Autel
  • Video: MISB/KLV, SRT

There are variations of these metadata formats, and some variations may not currently work. Please, inform us if you have such problems to info@i-conic.eu.

Running a project

Once a project is created or opened, one or more small panes (small resizable windows) are shown. These can be altered by clicking View.

Media

Now, use the Media buttons to run the 3D processing:

image-20230613101838681 Rewind = starts over from first frame/image [all your settings are kept]

image-20230613102156956 Play = runs 3D processing of all your data sequentially [changes to a Pause button when you play]

image-20230613102326922 Pause = stops processing

image-20230613102252009 Next = processes next frames or image pair

image-20230613102426645 Skip = skips x frames/images and processes those [x can be set in Settings]

image-20230613102454476 Redo = lets you reprocess the current frames or image pair after having changed parameters

While a video or image project runs, a low resolution 3D point cloud is generated from each frame. You may change the real-time resolution level in Settings. When you hit Pause, Next, Skip or Redo, a more detailed 3D model is generated.

File

image-20230613102545465 Save

Project. Allows you to save the project settings (using .icproj format). Default is to save in the output directory, which is C:\Users\[username]\AppData\Roaming\I-CONIC Studio\Projects. You may save it in another location with Save as...

Point Cloud. There are several save options for your point clouds:

  • Autosave point clouds. Will save point clouds with the interval set in Frame Interval.
  • Frame interval. This is the interval between point cloud alignments when Merge point cloud is selected, and frame pairs to skip when Merge point cloud is deselected. The default interval for Merge point cloud is calculated by the program, and set to 1 when deselected.
  • Save point cloud. Saves the currently displayed 3D point cloud to file.
  • Merge point clouds. Lets you start merging your point clouds. More information here.
  • Point cloud format. Lets you select which format you use for exporting the your point clouds; LAS or PCD. PCD is significantly faster to save, but can not handle coordinate reference systems.

Shapes. Saves the shapes in a file using *.wkt format.

I-CONIC. This option lets advanced users and developers to save images, point cloud, cameras, image and object coordinates for the current stereo pair to an .ibt file, for further processing. By default data are saved in binary format, but you can save them in CSV (Comma Separated Values) format through toggle off Use binary format*.

Set output directory. Your data is saved to files and folders in one and the same default output directory. If you want to change the default output directory, select Set output directory.

The other three File buttons will allow you to:

image-20230613102742712 Open your output directory. Default is C:\Users\[username]\AppData\Roaming\I-CONIC Studio\Projects

image-20230613102820763 Close the current project.

image-20230613102900240 Exit I-CONIC Studio.

Settings

In this Section, the basic Settings are described. Advanced Settings are found here.

image-20230613103557458 Undo/Redo

Lets you undo/redo your latest commands.

Options

image-20230613103518739 Offset

This is a quick way to increase or decrease the frame offset. More information here.

image-20230816162930390 Process


Basic process settings

Real-time resolution

To maintain real-time performance, a low resolution 3D point cloud is generated from each frame/image pair. Depending on the capacity of your computer GPU, you may want to change the real-time processing resolution. The lower number you select, the closer to full resolution you get (but also lose processing speed).

Export resolution

Whenever a point cloud is saved or you pause running a more detailed point cloud than the real-time resolution is generated. By default this export resolution is two, on a scale from one two six, where one is the highest resolution. If you do not require the highest resolution, you may change the export resolution so that processing goes faster and GPU memory requirements are reduced.

Frame process interval

Every n:th frame/image pairs are processed, other pairs are only displayed. Default is to process every pair = 1. Higher numbers will speed up the processing.

Frame offset

This parameter is used to select which video frames become a frame pair (aka stereo base). A value of 20 means, for example, that frame #1 and frame #21 are selected as the first frame pair. After that, #2 and #22 are processed and so on. You can manually change it by manually changing the Frame offset number. A lower number will decrease the base, which likely will increase the number of matched points, but also slightly decrease the height accuracy of each point. And the opposite for if you increase the base.

A suitable Frame offset can be calculated by the program. Check the Auto adjust offset box and then hit OK. This feature is by default turned off because at the beginning of a flight, the speed of the drone/platform is often not constant. For example, a drone may start by moving only vertically or hover of a place. A reliable frame offset can not be determined from under such circumstances.

If your video starts with a hovering drone before it starts flying forward, it is advised to auto-adjust offset from the start of the forward flight. See below for selecting start frame.

Start frame

Assuming that you run a project several times, and want to start at a particular frame, select Start frame. The next time you open the project, it will fast-forward to the selected start frame.

Last frame

Assuming that you run a project several times, and want to end at a particular frame, select Last frame. The next time you open the project, it will stop processing at the selected last frame.

You can find information about the Advanced setting here.

image-20230816163124076 Camera


Camera settings dialog

Camera Pre-set

  • Camera Pre-set. The camera chosen in the list is the camera used when calculating the 3D-data. A suitable camera is selected using metadata, e.g. EXIF, if available.
  • New. Here one can define your own camera definition. The camera will be added to the list.
  • Edit. Will show you the present camera parameters and allows you to make changes.
  • Delete. will delete the camera from the list.

In the menus for adding a new camera or editing a predefined camera, all values in the software is defined by defining a normalized camera with the largest camera dimension 1.0 (e.g. an image width of +/-0.5). For example, a 35 mm image with a 24 mm focal length should be set to 24/35 = 0.685714.

  • Name is the name shown in the image list
  • Make as the brand of the camera (optional)
  • Model is the camera model (optional)
  • Serial number of the camera (optional)
  • Focal length is the focal length expressed as a factor times image width (largest image side dimension)
  • Radial distortion A3 is the radial distortion defined below
  • Radial distortion A5 is the radial distortion defined below

Conflicting cameras

If a camera definition can be created from metadata, but there already is a camera definition with the same make, model and serial number, then a dialog is shown so that you can choose which one to use.

This may for example occur if you have calibrated the camera and perhaps defined a more accurate focal length or added radial distortion parameters. In such cases, you will likely override the metadata camera and select your defined camera instead.


Camera comparison dialog

Camera calibration file

If you have a camera calibration file, select it here. If your format is not supported, we appreciate your feedback to info@i-conic.eu.

DEM file

You are recommended to download a DEM file, even if you have terrain height data in your metadata, since they also include projection definitions and a global geoid model.

You can find information about the Advanced camera setting here.

image-20230816163203007 Point cloud


Point cloud settings dialog

Merge point clouds

Check this box if you want to merge your point clouds. It is done by aligning every n:th frame pair, where n is calculated by the program. If you desire to set n yourself, you can change that with Frame interval. When you hit Play, the program will start aligning the selected point clouds into a continuous point cloud. You may save the merged point clouds with Autosave point cloud.

Merging point clouds from a longer drone flight will eventually create more data than will fit on the GPU. When 25 million points have been generated, Studio will continue to create new point clouds but cache the oldest point clouds to file and delete thme from the GPU (and hence no longer show them). However, using the tree structure in the window Created Objects here, you may select and deselect any of the created point clouds for visualisation.

Point cloud merging works for video and still images, however merging of still images can be very sensitive to changes of direction in the drone flight. Please note, that you need to deselect Georeference (if preselected) before you can merge point clouds.

Problems with merging your point clouds? Find more information here.

Georeference

If you have provided metadata for georeferencing, then the point clouds are georeferenced by default. If you don't want to georeference your point clouds, then deselect this box.

Point cloud format

Lets you select which format you use for exporting the your point clouds; LAS or PCD. PCD is significantly faster to save, but can not handle coordinate reference systems.

Z-filter Min

Remove Z values in the point cloud below this threshold(0.1 per mille).

Z-filter Max

Remove Z values in the point cloud above this threshold(0.1 per mille).

Z-filter

When selected, this parameter will help you filter out noise in the Z axis. The default value is 10. For Z-value Min, this means that the lowest one (1) per mille of the points in the point cloud are removed, and opposite for Z-value Max [0.1 per mille for each step].

Median filter

When selected, the median filter will be active.

Autosave point cloud

If you check Autosave, it will save point clouds with the interval set in Frame Interval.

Frame interval

This parameter is used for interval between point cloud alignments when Merge point cloud is selected. The default interval for Merge point cloud is calculated by the program, and set to 1 when deselected.

image-20230816163253003 Merge Point clouds

This is a short-cut button to select/deselect merging of your point clouds. Details here.

image-20231027161701677 Auto adjust offset

When merging point clouds from video, it may be advantageous to automatically set the optimal frame offset. More info here. Toggle on/off with this button.

image-20230816163456279 Faster merging of point clouds

Due to large overlap, Studio do normally not use every point cloud for the final merging product, but every point cloud is used for background processing. If you want to speed up Studio processing further, you can select Fastest Merging. Since not all data is used for the processing, you might encounter accuracy issues. In such a case, start over and deselect Fastest Merging.

This is the same effect as described in Frame process interval above.

image-20230816170620962 User Plug-ins

Developers and advanced user can create their own plug-ins to I-CONIC Studio. You can find more information here.

Analyse

Some basic tools are currently available in Studio. You can easily export your created real-time point cloud to your favourite external analysis system.

Measure

You can only make measurements in the window First Frame/Image, and the drawn shapes can then be seen in the Single 3D Point Cloud window or the Merged Point Cloud window. If you have selected Merged here, then the shape will show up in the Merged Point Cloud window. If you then deselect Merged, you will see the shapes in the Single 3D Point Cloud window. Please note that you will lose the drawn shapes when you go from Merged to Single, or going back. You will get a warning and an option to save the shapes.

You can only make measurements within the overlap area of the latest frame/image pair. You can easily view this overlap area by toggle on View -> Frames/images -> Height colours here.

Information about your shapes is found in the window Created Objects. If your data isn't georeferenced or you haven't Set scale, then the distances, areas, etc will be estimated by Studio.

image-20230613104855250 The Navigate button helps you to move around when you are not measuring.

image-20230705145052538 The Select button allows you to select shapes for editing.

image-20230817142929487 The buttons for point, line, polygon, hole in polygon and colouring of shapes and are self-explanatory.

image-20230615165411491 Set scale makes it possible to create a correct scale even if you don't have metadata. Hit the button and draw a line between two known points. You will then see: image-20230705150320666 Write the length of the drawn line, and all measurements will be updated to the correct scale. Before you have set a scale, Studio makes a qualified estimation.

Edit

image-20230817143020480 You may also delete or save the shapes you have created. Shapes are saved in a file using *.wkt format.

External analysis

image-20230816174130015 When hitting this button you may select your favourite External application of your choice.

image-20230613105739236 By clicking this button you can immediately open the current point cloud in your favourite external analysis software application.

image-20230613105809551Point cloud format. Here you may select which format you use for exporting the current point cloud; LAS or PCD. PCD is significantly faster to save, but can not handle coordinate reference systems.

image-20230613102742712 Here you can open an external point cloud file.

View

With the first seven Windows buttons you may open/close different windows:

image-20230613105933555 1st Frame/image.

This is the first frame or image used in the frame/image pair. In this window you can also draw shapes with the Measurement tools.

image-20230613110006621 2nd Frame/image.

This is the second frame or image used in the frame/image pair.

image-20230613110108201 Real-time point clouds.

This window will show all the point clouds that are processed in real-time.

image-20230613110617630 Merged point clouds.

This window will show all the aligned point clouds. You may select/deselect individual point clouds in the tree structure shown in Created Objects. This can, for example, be helpful if some of the aligned point clouds are corrupt.

image-20230613110659769 Map.

If you have metadata for interior and exterior orientation elements, the map view will be displayed by default and a trajectory of the project is drawn when project is loaded. Thumbnails footprints are added to the map while processing.

You may select between different background maps in the upper right corner.

image-20230613110734979 Stereo image.

This window allows you to view the current frame/image pair in 3D stereo. You need red-cyan anaglyph stereo glasses for viewing. If run a 2D video, you'll see a 3D stereo video in this window.

image-20230613110815562 Created objects.

Two types of created objects are shown in this window:

  • Tree structure with all the point clouds in the Merged point cloud. You may select/deselect individual point clouds, which can be helpful if some of the aligned point clouds are corrupt.
  • Information about each of the shapes you created by the Measurement tool.

You can right click any point cloud or shape tree node to e.g. delete or save the corresponding object. You can right click the root node for all point clouds or shapes to delete or save all selected objects.

image-20230613110854328 Layout

Click Layout and you will get the default layout for I-CONIC Studio. The arrow next to Layout allows you to Load saved layouts (including default), Save a a layout you have created, or Delete saved layouts.

View

image-20230618103719527 Toggle on/off Full screen.

image-20230618103934476 Toggle on/off Borders around the windows.

Frames/images

These buttons can affect both the windows 1st Frame/image and 2nd Frame/image, and help you to quickly

image-20230613111207762 Zoom in/out.

image-20230613111355944 Fit frame/image to the window.

image-20230613111425725 Toggle on/off Height colours.

Point cloud

These buttons affect both the windows Real-time point clouds and Merged point clouds. Here you can

image-20230705151330184Recentre the model in the window

image-20230613111823566 Toggle on/off 3D stereo viewing of the point clouds. You need red-cyan anaglyph stereo glasses for viewing.

image-20230615165411491 Change the stereo effect.

image-20230613111910551 Toggle on/off Height colours.

image-20230705151723121 Change the point size in your 3D model for optimal viewing.

Info

image-20230613111957544 Metadata. Toggle on/off window to show your metadata, which is extracted from the images while processing. The view shows parameters for each of the two images in a summarizing a tree with all extracted parameters.

image-20230613112025907 Log. Toggle on/off window to show processing statistics.

image-20230613112053688 Help. Here you can find more information about your system, downloading of a new version, as well as a link to this manual.

Advanced Settings

Process

Feature threshold

Edges are extracted in each video frame with sub pixel accuracy. The edge threshold defines the minimum strength of edges. Lowering this threshold will thus create more edges (but in general also more noise).

Search space X/Y

Search space defines the neighbourhood of pixels to examine for matches for X axis and Y axis, respectively. For most parts this is automatically set, but this threshold defines the search space at each pyramid level when a stable sensor orientation has been found. A larger search space is automatically used when the solution is unstable.

Image registration

A course image registration is performed at project start and whenever matching is automatically identified as unsuccessful. If there are certain metadata, such as camera position and orientation, then a module for metadata based image registration is automatically selected. If such metadata are missing, then by default a less accurate translation based image registration is selected. If your video or still images contain sharp changes in orientation (primarily yaw angle), then you are recommended to switch to the translation-rotation image registration, which is slightly slower, but more robust to changes in roation You can always select the translation or translation-rotation based image registrations although there are metadata, but not the opposite.

Match window

The size of the square window that is used as neighbourhood for each image point in template matching. In general, a larger window will result in more robust results, but details may be lost and processing becomes slower.

Match threshold

The match threshold defines the minimum similarity between matched edges for the match to be accepted and used for sensor modelling. Lowering this threshold will accept more matches (and increase the risk to include more incorrect matches).

Camera

Sensor type

Perspective is the default and the best solution for most projects. In some cases, for example if you have a large radial distortion and you cannot provide information in the Camera Pre-set menu, then projective might help.

Rectify

Initially, the point cloud is defined in a 3D projective space, where e.g. coordinate axes not necessarily are perpendicular. This 3D projective space is upgraded to a 3D Euclidian space when the Rectify button is checked. This process includes a simplified self calibration, which is not guaranteed to succeed. However, you are recommended to keep the Rectify button checked unless you find the resulting 3D space distorted in which case it may help to uncheck rectification.

Compute 3D accuracy

When this is checked, the plane and height accuracy of the 3D coordinates are estimated. They are not used internally, but will be exported to ibt files. To avoid unnecessary processing it is recommended to keep this unchecked unless you want to export plane and height accuracies to file.

Logging

Errors, warnings and important messages are shown in dialogs.

Statuses like processed number of frames per second (fps) are shown in the status bar at the bottom left of I-CONIC Studio.

When I-CONIC Studio is running, a log file is saved in C:\Users[username]\AppData\Roaming\I-CONIC Studio\log. If you run into unexpected behaviour, please open the log file and check the last few lines. Maybe you will find information about what went wrong so that you can correct the error. If not, and if this documentation does not inform you on what went wrong, then please always submit your latest log file to your support issue. Log files are deleted when they are 24 hours old.

Known issues and potential mitigation

This is a Free Trial of I-CONIC Studio. You might find issues that we haven't found, then please message us at info@i-conic.eu.

I-CONIC Studio crashes

  • If you haven't updated the GPU drivers, it might make Studio crash. You can find the available drivers here: https://www.nvidia.com/download/index.aspx. Information about your system is found here.
  • If your GPU memory is insufficient, you may reduce the export resolution which will reduce the GPU memory requirements
  • If you are starting a project with too large images/video frames, Studio might crash. The acceptable image size depends on your GPU.

Measurements

  • When you have reached End-of-file, it is no longer possible to make measurement.
  • After Rewind, some measurements could be slightly off its correct position. (This will be solved in a new release.)

Merged point clouds

  • Point cloud merging works for video and still images, however merging of still images can be very sensitive to changes of direction in the drone flight. Merging of still images can only be done when turning Georeference off in Settings->Point Cloud.
  • If you have a large radial distortion in your camera (the wider the lens, the greater the distortion) and you haven't added such information in the camera pre-set, then you will notice errors (noise) in the merged point cloud.

Low-quality point clouds

  • Low-quality or non-existent 3D models may depend on the input data, for example:

  • the two images/frames in the pair do not overlap at all.

  • footage taken when the drone is hovering over the same area. Models begin to be created when the drone starts flying.

  • the images/frames in the pair are rotated in relation to each other (e.g. when the drone is turning).

  • there is very little contrast between, or inside, objects in the images/frames (e.g. areas with snow or water).

  • there are objects moving between the take of the first and second image/frame in the pair. Moving objects might disappear or disturb the model.

  • there is very large distance between the closest and farthest objects, e.g. when the images/frames are taken towards the horizon

  • Sometimes the quality of point clouds from video can be improved by changing frame offset here, or by using the advanced processing commands here.

Changelog

Version 1.0.6 (2024-01-23)

  • Added new windowing information for identifying the files that are being processed
  • Added a new settings menu to the main menu
  • Added a new side by side stereo view
  • Added functionality for defining a last frame to be processed
  • Added a coordinate reference system (CRS) menu for choosing lat/long and vertical coordinate systems
  • Bug fix: Fixed an issue when saving point clouds from the tree view
  • Bug fix: Fixed an issue where the camera selection screen would sometimes not display the right selected camera.

Version 1.0.5 (2023-10-27)

  • Can now read SRT metadata
  • External 3D point clouds can be imported
  • Does not ender point clouds that fails to merge
  • Skip video frames when the overlap is too small to generate point clouds
  • Improved rendering trajectory when project is loaded
  • Processing first point cloud when project is loaded
  • Toggle button for auto offset
  • Progress bar when loading project
  • Resetting merge when pyramid level has changed

Version 1.0.4 (2023-10-03)

  • Significantly improved 3D generation for imagery with large changes in rotation (primarily yaw), using a new translation/rotation image registration. Since this image registration is slower than the original translation image registration, it is however not used by default.
  • Processing can be terminated at a lower resolution than the full image resolution
  • Vertical datum of exported point clouds now refers to geoid height EGM08
  • Improved feedback when there are conflicting camera definitions.
  • Licenses for Studio can now be retrieved online
  • Auto adjust offset is now turned off by default because it is normally more useful after processing a while so that the speed of the drone/platform has stabilized
  • Point cloud merging is turned off by default because the primary focus is to create instant 3D models of your current view. While merging will be further improved in later releases, we still enable merging, but you will have to turn it on.
  • Bug fix: Camera definitions could be saved twice in the camera definition file
  • Bug fix: Logging to file is now activated by default
  • Bug fix: Corrected link in Studio to these release notes

Version 1.0.3 (2023-08-30)

  • Improved 3D coordinate calculation from image measurements
  • Add Coordinate Reference System (CRS) to saved pointclouds
  • Bug fix: Measurements are no longer off after rewinding video

Version 1.0.2 (2023-08-18)