It’s a Snap!

Take a photograph and create a 3D

Model
Dominique Pouliquen – Autodesk, Inc.

AC4056

Yes, all it takes is a few properly taken photos to create a 3D model with Autodesk Labs Autodesk
123D Catch which is now known as Autodesk 123D Catch.

With this technology you can create a 3D model using only photos taken using a consumer digital
camera. You can capture your house, a large building, a sculpture, even your own head in minutes.
Autodesk 123D Catch generates not only a point cloud but a surface mesh textured with materials
from the photos. You can export the 3D model to many file formats or even export a video. We will
show you how to use Autodesk 123D Catch and helpful tips and tricks to get the best resulting
3D model. We will also cover what can be done with the 3D generated data including using it in
Autodesk products like AutoCAD, 3ds Max and Revit.

Bring a digital camera, after this session you will be ready to begin capturing objects around you in
3D.

Learning Objectives

At the end of this class, you will be able to:

• You will understand what Autodesk 123D Catch is and how the technology works.
• Understand many of the uses and customer examples of Autodesk 123D Catch. Learning
objective 3
• Know how to take photos using your camera to generate the best resulting 3D model.
• Understand how to use the Autodesk 123D Catch created 3D model data in other applications.

About the Speaker:

Dominique Pouliquen is Product Manager for reality capture technologies in Autodesk Labs. He
is a member of the Computer Vision & Imaging (CVI) team of Autodesk based in Sophia Antipolis,
France. Autodesk bought REALVIZ in May 2008, a company that Mr. Pouliquen co-founded in March
1998 from a technology transfer with INRIA (National Research Institute for Computer Sciences and
Automatics). He was CEO of this company for ten years, developing its image-based content creation
software business in the special effects and digital imaging markets all over the world. Mr. Pouliquen
has spent his career in the computer graphics industry after graduating as an Engineer from Ecole
Superieure d'Electronique de l'Ouest, Angers, France in 1986.
dominique.pouliquen@autodesk.com

You can download 123D Catch Beta free from http://

www.123dapp.com/catch

It’s a Snap! Take a photograph and create a 3D Model - Page 1

Step 1 – Shooting the right Images
Your photos must be located on your disk for further use in 123D Catch Desktop
Beta or in an Autodesk application. During the Stitching, they will be uploaded
temporarily to the cloud for the computation of the Photo Scene.

More video instructions on: http://www.123dapp.com/catch

It’s a Snap! Take a photograph and create a 3D Model - Page 2

Step 2 – Creation of a Photo Scene
2.1 – Global workflow

Photo Scenes are computed in the cloud, which means your Images will be uploaded
to a server for processing, and you will get a Photo Scene file back once the
computation is over.

The returned Photo Scene includes a photo-textured 3D mesh, and a 3D

camera for each original picture that could be stitched automatically.

Default Photo Scene Name

The Photo Scene file has a “.3dp” extension. This file contains the necessary
links to retrieve your data from either your disk or the servers. This file does
not contain any image or 3d data. It is a light file that you can easily send to
others via e-mail, if you want to share your projects. Note that a default name is
automatically created to include the date and the time when it was computed, as
follows: “Scene_yyyy_mm_dd_hr_mn_sc.3dp”. You can change this name and set yours.

It’s a Snap! Take a photograph and create a 3D Model - Page 3

2.2 – Creating a Photo Scene / a 3D model

- Run “123D Catch.exe” to start 123D Catch Desktop Beta wizard. If you are
connected to the Internet, the following window will pop-up:

- Click on “Create Photo Scene From Your Images” to start the process. This
will open the image selection window.
- Browse and select your photos. Click on “Open” when finished. Click
on “Compute Photo Scene” to start the stitching process.

It’s a Snap! Take a photograph and create a 3D Model - Page 4

Wait or E-mail
Once your Images are uploaded, processing starts in the cloud. You then have the
choice to:
- Either wait for the result and stay connected:
o Once the computation is over, the Photo Scene will appear in a 3D
viewer automatically with no intervention.
o The .3dp file will automatically be saved in the same folder as your
Images.
- Or quit the application and get the result later, since the processing may take
a while depending on the number of Images, the number of projects being
processed, and the number of CPUs allocated in the cloud at that time. In this
case, you are asked to provide your e-mail address, where a message will be
sent once the computation is done. You can get the result by either opening
or saving your Photo Scene.

It’s a Snap! Take a photograph and create a 3D Model - Page 5

 - The “Photo Scene Name” will allow you to enter a name for your project, so
that you can easily identify it in the e-mail giving the link to download the
result once the computation is completed. This can be useful if you have
launched several projects on the cloud.
- Click on one of the options.
- The Images will then start getting uploaded to the server for computation.

Receiving an e-mail once the computation is finished

- Once the computation is over, you will receive an e-mail from Autodesk
asking to clik on a link to download your 3dp file.

It’s a Snap! Take a photograph and create a 3D Model - Page 6

2.3 – Opening a Photo Scene and interacting with the 3D model and the Photo
Scene
Opening a Photo Scene
- Run “123D Catch.exe” to start 123D Catch Desktop Beta wizard. The
following window will pop-up:

- Click on “Open an Existing Photo Scene” and browse to select your 3dp file.
- Note that if your 3dp file has been received from another user who was
willing to share his project with you, the software will automatically retrieve
the Source Photos and the 3D model from Autodesk servers.

Selecting the interface layout

123D Catch Desktop Beta main components are:
- The menu bar
- The ribbon : composed of tool buttons and commands; Includes a sub-ribbon
- The 3d workspace : the 3d view where all objects are drawn
o 3D Navigation manipulators
o Photo Navigation Manipulators (Thumbnail View, Photo Space
Navigator)

Menu Bar

Tool
It’s a Snap! Take a photograph and create a 3D Model - Page 7 Bar
A sub-tool bar is also available for more display options:

Selecting the full screen layout, or coming back to the default layout, can be set from
the menu bar by: View > Toggle Full Screen, or by pressing Alt + Enter.

It’s a Snap! Take a photograph and create a 3D Model - Page 8

Viewing the stitching results
The stitching results are available from the menu bar: Help > Project Information, or
by pressing the I key.

A pop-up window will then appear showing the stitching results:

Navigating in the Photo Scene

123D Catch Desktop Beta provides 2 different navigation modes:
• Free “3D navigation” mode = Navigating and exploring the 3D space.
• “Photo Lock” mode = Looking through a Calibrated Camera in a constrained
navigation locked to an image, and navigating from Calibrated Camera to
Calibrated Camera.

Switching from the “3D navigation” mode to the “photo lock” mode
This is performed either by:
- Clicking on a thumbnail.
- Pressing the space bar on the keyboard.
- Right clicking in the 3D workspace and selecting “Photo Lock.”
- Clicking on one of the white arrows in the lower right corner of the 3D
workspace:

The directions to the available adjacent photos are drawn filled, and can be selected.
The disabled directions (no available adjacent photo) are drawn in wireframe.

Switching from the “photo lock” mode to the “3D navigation” mode
This is performed either by:
- Pressing the space bar on the keyboard.

It’s a Snap! Take a photograph and create a 3D Model - Page 9

 - Pressing Alt + Double clicking on a thumbnail.

Switching from one photo to the next in the “photo lock” mode
This is performed either by:
- Clicking on a thumbnail.
- Pressing Alt + Left or Right keyboard arrow.
- Orbiting by pressing Alt + Right click in the 3D workspace: this mode will
temporarily use a 3D camera motion to move through the various photos.
The lock on the nearest camera will happen when releasing the right click.
To get this behavior, make sure that “Lock during Photo Orbit” is set in the
Preferences -> General Settings (see Annex 1).
- Clicking on one of the white arrows in the lower right corner of the 3D
workspace.

Navigation tools in 3D
The free 3D navigation tools are the following:

Tool Ribbon Shortcut

3D PAN : Middle click

ORBIT + : Alt + Left click

+ : Ctrl + Alt + Left click

ZOOM Or
: Wheel

ROLL
(= camera rotation +
around its viewing + : Alt + Right click
Rotation
axis) in the 3D
window

Free 3D navigation tools are also available from the menu bar: View > Navigation.

“Bird’s Eye View” mode

The Bird’s Eye View is an automatically created viewing camera position that shows
the entire Photo Scene, i.e. the 3D mesh and the Calibrated Cameras displayed with
their 3D icons. When switching from a “Photo Lock” mode to the Bird’s Eye View,
the viewing camera axis keeps the same direction than the Calibrated Camera on
which it was locked just before.

It’s a Snap! Take a photograph and create a 3D Model - Page 10

The Bird’s Eye View can be toggled through the main contextual menu (right click in
the 3D window):

It can also be toggled through the menu bar: View > Photo Lock or through the
<Space> key.

Grid display
A grid can be displayed on the computed default XOY plane.
- The size of the grid will be relative to the scene (bounding box).
- You can add/remove the grid by clicking on « Grid » in the sub-ribbon.
- You can always remove the grid in the Preferences

3D mesh display
The default display is a photo-textured 3D mesh. Display options for the mesh are
accessible in the sub-ribbon. You can select either the wireframe or the textured
mode. A slider will set the transparency level over the background. This is useful to
check the superimposition of the 3D mesh over the cameras in the Photo Lock mode.
Wrong or badly stitched cameras appear then clearly.

With the 2-sided Surfaces option toggled, the triangles of the 3D mesh will be seen
in all directions, even from the back.

It’s a Snap! Take a photograph and create a 3D Model - Page 11

Navigation Tools in the “photo lock” mode
The navigation tools available in the “lock on camera” mode allow moving in the 2D
photo space. They are the following:

Ribbo
Tool Shortcut
n

+ : Alt + pressing down the

2D PAN
wheel

ORBIT (this tool is used to

move to another photo in the
lock on camera mode – See
above)
Needs to set + : Alt + Left click
“Lock during photo orbit”
in Preferences > General
Settings
+ : Ctrl + Alt + Left click
ZOOM Or
: Wheel

2.4 – Requesting other mesh densities

123D Catch Desktop Beta offers the capability to create 3D models in 3 different
preset densities. The first stitch will automatically produce a 3D model in “standard”
density. You can ask for another mesh density (mobile or maximum) by clicking on:

-
Click on to make the following pop-up window appear:

It’s a Snap! Take a photograph and create a 3D Model - Page 12

- Click on the desired mesh quality.
- The Photo Scene will then be uploaded on Autodesk servers for computation.
Note that the Source Photos are NOT be uploaded a second time. The
automatic mesh creation process may take several minutes depending on
the size of your project, the number of projects in the queue of Autodesk
servers, and the quality that you requested. Therefore, you can either wait
or request an e-mail to be sent once the computation is over. The e-mail is
recommended if you requested a mesh with maximum quality.

“Mobile” quality mesh

It’s a Snap! Take a photograph and create a 3D Model - Page 13

 “Standard” quality mesh

“Maximum” quality mesh

2.5 – Manual Stitch

The manual stitching process consists in adding the photos that could not be
stitched automatically in your project, by manually connecting such missing photos
to the ones already stitched. With 123D Catch Desktop Beta, you can now manually
stitch several unstitched photos in a single pass before re-submitting your Photo
Scene to the Autodesk servers: each photo that you have been manually stitching
is pre-computed locally and pre-positioned at its right location. Therefore, each
manually stitched photo can be used as any other stitched photo as a reference
for connecting other unstitched photos. This is particularly useful when you need
to complete the complete tour of a scene (a building, an object…etc.) with several
pictures that requires a manual stitch.

It’s a Snap! Take a photograph and create a 3D Model - Page 14

Here is an example below. Several pictures could not be stitched automatically, and
they appear darker in the thumbnail viewer.

- Double click in the thumbnail picture of the first photo that you want to
manually stitch, or right click in it and select “Manual Stitch.”

- This will pop-up the following window:

It’s a Snap! Take a photograph and create a 3D Model - Page 15

 - The process is the following:
o The picture to be stitched manually is displayed on the left hand side.
You can browse the unstitched pictures left or right using the arrows
on the sides of the displayed photo.
o On the right hand-side, you can select and view 2 pictures among all
the stitched pictures. You can browse the stitched pictures left or right
using the arrows on the sides of the displayed photos.
o You must create at least 4 connections between the non-stitched
picture on the left, and at least 2 stitched ones on the right.
o Start by selecting a feature in the non-stitched picture by right clicking
on top of the image. Make sure to select an area that is visible in at
least 2 stitched images.

Manually stitching one picture

- Follow steps 1 to 6 below.

It’s a Snap! Take a photograph and create a 3D Model - Page 16

 Step 1 – Select a feature (well defined, highly contrasted point)
in the left view (non-stitched picture). A left click in the image will pop-up a
magnifier window to allow for a sub-pixel positioning of the point.

Step 2 – Select the same feature in the right views (stitched pictures).

It’s a Snap! Take a photograph and create a 3D Model - Page 17

Step 3 – If the “Suggest Points” button is set at bottom left of the window, the software
will automatically propose a matching feature in the third view as below. Proposed
features are marked with a yellow square. You need to click in it to validate or readjust
the location of the point in this 3rd view.

Step 3.1 – If there is no proposed feature, you must click at the right location to
complete the manual matching process on the third view as well.

Step 4 - A first connection is now in place.

It’s a Snap! Take a photograph and create a 3D Model - Page 18

 Step 4.1 - If the connection point remains yellow in all 3 windows as above, it means
that there is not enough parallax between the 2 selected stitched photos on the right,
so you will need to reinforce this point by selecting it again in a 4th view showing more
parallax as in the example below.

Step 5 – Repeat steps 1 to 4 above for 3 more points at

It’s a Snap! Take a photograph and create a 3D Model - Page 19

 least.

- Once 4 connections have been made with all manual points in “green”,
the picture that you just manually stitched has been pre-processed at its
right location. It will be used as a reference for further manual stitch. The
following message appears:

This added photo then appears in the thumbnail viewer as follows:

Manually stitching more unstitched photos

- You can then manually stitch all the other unstitched pictures if necessary,
by repeating steps 1 to 6 above, before submitting your project to Project
Photofly servers.
- When you are done connecting all the unstitched pictures, click on “Submit”.
The information is then sent to the server for a new computation of the Photo

It’s a Snap! Take a photograph and create a 3D Model - Page 20

 Scene. As explained in section 2.2, you may wait for the result, or “Quit and
Receive the Photo Scene Later.”

Managing the manual stitch issues: Case 1 – Wrong point

- If your manual point is not properly located in one of the 2 stitched photos, it
is then marked with a red square as below.

- You will then be prompted the following message as well:

- You will then need to re-position this point at its right location.

Managing the manual stitch issues: Case 2 – Wrong stitched camera

- It may happen that one stitched camera (appearing in the right windows as
a reference one for connecting points) was badly stitched by the automatic
engine, and you did not see it when checking for wrong cameras as explained
in section 2.4.
- In this case, all the points clicked in this picture will appear red, as in the
example below.

It’s a Snap! Take a photograph and create a 3D Model - Page 21

- If you validate each one of these red markers in this picture (by clicking
successfully in each red square), the following message appears:

- Clicking on “OK” will then re-position the corresponding camera at its best
location. You will then need to re-submit the Photo Scene to Project Photofly
servers to fully validate it.

It’s a Snap! Take a photograph and create a 3D Model - Page 22

It’s a Snap! Take a photograph and create a 3D Model - Page 23
2.6 – Adding Images
You can add one or more Images at any time in your Photo Scene. To do this, simply
select Edit > Add Images and Stitch… from the menu bar.

- This will open the Image selection window.

- Once the photos have been selected, click on “Open”. This opens the following
pop-up window:

- Clicking on “Process Scene” will start the automatic stitching process,

including the upload of the selected photo(s) to the server.
- Clicking on “Return to Scene” will add the selected images in the thumbnail
view as unstitched. You will then need to manually stitch them as explained
in section 3.12 – Case 1 here above.

2.7 – Sharing projects

Sending a 3dp file
You can easily share your projects with others by simply sending your 3dp file as an
attachment in an e-mail. The person receiving this 3dp file will need to install 123D
Catch Desktop Beta to open it.

The 3dp file includes some Photo Scene data and references others on your disk
and/or the Project Photofly servers:
- Data included in the 3dp file:
o Camera information
o Reference Points, Lines, measurements, scale, WCS, if added by the
user
o Manual stitch data
- Data NOT included in the 3dp file, but referenced by it:
o Source Photos
o 3D mesh (draft or any other density)
o Points Clouds

It’s a Snap! Take a photograph and create a 3D Model - Page 24

When the person will receive your 3dp file, 123D Catch Desktop Beta will look for
the corresponding referenced data on Project Photofly servers. Note that this may
take some time to download all this data since images and mesh can be heavy.

Synchronizing your project

If you want to share your project in its current state, you must first synchronize it
with the one stored on Project Photofly servers. Synchronization occurs when:
- A scene is calculated/processed (transparent for the user).
- You save a file and the corresponding preference is turned on (default:
turned off).
-
You click on in the tool bar.

See Annex 1 to turn ‘Sync on Save’ on or off. This preference is set to OFF by default
(= no synchronization on save).

Sending the Photo Scene by e-mail

Click on to send an-mail inlcuding the necessary links to download your

project. This will automatically create an e-mail with the following text included:

“Hi,

I’d like to share my 123D Catch project with you!

Here is a link to download the project: <http://cvi-data-photofly.labs.autodesk.com/
MediumRes.3dp>
123D Catch is a free application from Autodesk which lets you turn photographs into 3D
models. You can get your own copy here: http://www.123dapp.com/catch »

2.8 – Deleting your project from Autodesk servers

You can decide to delete all your project data from Autodesk servers at any time. To
do this, simply select Edit > Delete Project Data from Server from the menu bar.

Note that if your images are used by other older projects, they will not be deleted
from the servers until you have also deleted these older projects.

It’s a Snap! Take a photograph and create a 3D Model - Page 25

Step 3 – Edition of the Photo Scene
3.1 – Cleaning the 3D mesh

At any stage in the process, it might be useful to clean the 3D mesh, either for a
better viewing experience, or to select a specific area of the draft mesh before
requesting a mesh refinement.

Cleaning the 3D mesh is performed by:

- First selecting the desired triangles, using the rectangular selection or the
lasso selection tools available in 123D Catch Desktop Beta.

- The display should be set to “3D Navigation” mode with “Select All (No
Cameras)” or “Select Faces” to allow for an easier selection.
- Using one of the above tools, you can then select the desired triangles, as in
the example below. Once selected, they will appear in red (default viewing
option).

- You can then delete these triangles by selecting Edit > Delete from the menu
bar, or by right clicking in the 3D workspace and selecting “Delete”.

3.2 – Creating Reference Points

Why Reference points?
You may need to create 3D points at locations of interest, for various purposes:
- Setting the World Coordinate System (WCS): the WCS requires a well-located
3D point that will become the origin, and at least 2 pairs of 3D points (the
origin may be one of these points) that will define the X, the Y or the Z axis.
- Setting the scale of the Photo Scene: the scale will be set by assigning a
known distance between two 3D points where the distance indicates real-
world measurements.
- Creating measurements: measurements will be defined between two 3D

It’s a Snap! Take a photograph and create a 3D Model - Page 26

points, or between one 3D point and along one of the WCS axis.

It’s a Snap! Take a photograph and create a 3D Model - Page 27

 - Creating lines and polylines representing edges of the scene.
- Image-based modeling: Once in a 3D modeling package such as AutoCAD,
you can use the defined points and lines as a basis to create more complex
shapes. For instance, by snapping 3D primitives on well-located 3D points
such as corners of a building you can create the building envelope.

Image-based modeling step 1

Snapping the basic shape to 3D points
at the corner of a building

Image-based modeling step 2

Adding modeling details by attaching
more primitives to the basic shape
and aligning them with the Background Pictures

Creating a Reference Point

To create a Reference Point, follow the steps below:
- Select your viewpoint: The preferred viewpoint is the one right in front of the
3D point that you want to create.
- Set your display mode.
o The mandatory visualization mode for creating Reference Points
is “Photo Lock”.
o You should then remove the 3D mesh from the display since the 3D
points will be created with the pictures as references, not the mesh.
Uncheck the “Show Mesh” box in the sub-ribbon.
-
Click on in the tool bar, then select “Create Reference Point” from the
list.
- You can also click on Edit > Create Reference Point from the menu bar. Your
cursor then appears in the 3D window.
- Click on the desired feature in the selected image. Features to look for
should be highly contrasted and should correspond to an easily identifiable

It’s a Snap! Take a photograph and create a 3D Model - Page 28

 element of your scene (corners for example). When clicking on the feature, a
magnifier will pop-up, to allow for an accurate placement.

- Clicking on a feature will automatically run an automatic matching process,

and a 3D position will automatically be computed for this feature once the
same feature has been found in at least one other Calibrated Image, allowing
the triangulation process to occur.
- Once the 3D point has been created, it will appear with a red pin on the
image. It has now become a 3D Reference Point.

- The thumbnails will also show:

o The image in which the feature was initially identified, this thumbnail
being highlighted by a green pin.
o Green dots in the other images that give the visual indication that
your 3D point is properly located.

- At this stage, you are done with this 3D Reference Point, and you can create

It’s a Snap! Take a photograph and create a 3D Model - Page 29

 additional ones once the “Create Reference Point” tool is active.

Adjusting a Reference Point

In some cases, you may find that the 3D Reference Point created from one image is
not properly located in other images. You can then fine tune the manual placement
of the 3D point, using another image, at least, as a reference.
- First select another camera from which the initially selected feature can
be seen as well, but from a substantially different angle: the closer to 90°
angle between 2 viewpoints, the better for the triangulation process. As an
example, selecting the closest neighbor to the first image is typically not
appropriate.
- In this second image, the 3D point will be displayed as a red dot together
with a “guiding line.” This guiding line displays the optical axis on which the
3D point should be located, given its location in the first image. Make sure
that the “Display Guiding Line” is toggled in the sub-tool bar.

Example of 3D point & Guiding Line

in a second image

-
The cursor must appear as , which indicates that you are in the
adjustment mode. Clicking on the picture will then pop-up the magnifier
again. You can then easily adjust the placement of the point along this
guiding line, or close to it. Once you release the click, the triangulation
process is run again to create the 3D Reference Point.
- You may repeat this process for a 3rd and a 4th image to increase the accuracy
of your Reference Point.
- Every picture that was used to manually fine tune the placement of your
Reference Point will be tagged with a green pin in the Thumbnail View as in
the example below:

It’s a Snap! Take a photograph and create a 3D Model - Page 30

Labeling a Reference Point
Labeling a Reference Point may be useful in a design process. To create a label,
follow the steps below:
- Set your visualization mode as “Lock on Camera.”
- Make sure that “Labels” is toggled in the sub-tool bar.

- Select one Reference Point by clicking on it in the 3D workspace.

- Right click to get the contextual menu and select “Rename Point….” You can
also select Edit > Rename… from the Menu Bar.
- A small pop-up window appears, where you can type the name of this
Reference Point, as in the example below:

Deleting a Reference Point

To delete a Reference Point, follow the steps below:
- Select one Reference Point by clicking on it in the 3D workspace.
- Right click to get the contextual menu and select “Delete Point.” You can also
select Edit > Delete from the Menu Bar.

3.3– Setting a World Coordinate System

A default World Coordinate System (WCS) is set by the Stitching engine, with Z
up. You may find it useful to re-locate this WCS so that it better meets your design
criteria. This can be very useful to locate the WCS at the corner of a building as in
the example below:

It’s a Snap! Take a photograph and create a 3D Model - Page 31

To create your own WCS, follow the steps below:
- Select your viewpoint: The preferred viewpoint is the one that allows seeing
the entire set of Reference Points used for the placement of your WCS.
- Set your display mode.
o The mandatory visualization mode for creating Reference Points
is “Photo Lock”.
o You should then remove the 3D mesh from the display since the WCS
will be created with the pictures as references, not the mesh. Uncheck
the “Show Mesh” box in the sub-ribbon.
-
Click on in the Tool Bar and select “Define World Coordinate System” to
enter the WCS creation mode. You can also click on: Edit > Define World Space
in the Menu Bar. The following WCS icon appears somewhere in your 3D
workspace:

It’s a Snap! Take a photograph and create a 3D Model - Page 32

- Place the origin first, by clicking on the corresponding dot in the WCS icon,
and attaching it to one Reference Point.
- Place a first axis then, by clicking on one or both dots defining each axis in the
WCS icon. The selected dot appears in yellow.
o When clicking on the dot opposite to the origin, your selected axis will
move around the origin. You then virtually attach this dot to another
Reference Point that is located along this axis from the origin. In the
example below, the Z axis was attached to a Reference Point exactly
vertical located vertical from the selected origin:

o When clicking on the dot close to the origin, you may locate your
selected axis in a completely different area from the one used
previously. The second dot for the same axis must be moved as well
accordingly, as in the sequence below:

Step 1 – Select 1st dot of X axis Step 2 – Move it to its right place

Step 3 – Attach it to its Step 4 – Attach 2nd dot to 2nd

Reference Point Reference Point

It’s a Snap! Take a photograph and create a 3D Model - Page 33

 o You can use either one of these two techniques for setting two axes.
The 3rd axis will be deduced automatically.

3.4 - Setting a Reference Distance

You can define a Reference Distance to scale your Photo Scene, by setting a known
measurement between 2 Reference Points. Setting the scale will allow to make
measurements relative to this scale.

Note that the Reference Distance is unit less. You may set a Reference Distance in
whatever unit and then measure in this same unit. When importing this data in
AutoCAD for example, you will need to define your unit.

To define the reference distance, follow the steps below:

- Select a viewpoint that allows seeing the 2 Reference Points used for the
setting of your Reference Distance.
- Set your display mode.
o The mandatory visualization mode for creating Reference Points
is “Photo Lock”.
o You should then remove the 3D mesh from the display since the
Reference Distance will be created with the pictures as references, not
the mesh. Uncheck the “Show Mesh” box in the sub-ribbon.

-
Click on in the Tool Bar and select “Define Reference Distance” to enter
the reference distance creation mode. You can also click on: Edit > Define
Reference Distance in the Menu Bar. You are then given the opportunity to
attach the reference distance (displayed as a blue arrow) to 2 Reference
Points as in the example below:

It’s a Snap! Take a photograph and create a 3D Model - Page 34

 - Once you place the second end of the arrow to the second Reference Point,
a pop-up windows appears, in which you can enter the value of the real
measurement:

- Clicking on “OK” will scale the entire Photo Scene.

3.5 – Making Measurements

You can measure distances between 2 Reference Points, or between one Reference
Point and along one axis of the WCS. Distances are relative to the Reference Distance
that you previously set up.

To create a measurement, follow the steps below:

- Select a viewpoint that allows seeing the 2 Reference Points used for the
measurement.
- Set your display mode.
o The mandatory visualization mode for creating Reference Points
is “Photo Lock”.
o You should then remove the 3D mesh from the display since the 3D
measurements will be created with the pictures as references, not the
mesh. Uncheck the “Show Mesh” box in the sub-ribbon.

-
Click on in the Tool Bar and select “Create Distance Measure” to enter
the measurement creation mode.
- To measure the distance between 2 Reference Points, click the first end of the
ruler on the first point, and the second end on the second point. The distance
will then appear as shown below:

It’s a Snap! Take a photograph and create a 3D Model - Page 35

- To measure the distance between one Reference Point and along one axis of
the WCS, click the first end of the ruler on the first point, and extend the ruler
as much as needed along the selected axis that will automatically appear in
red/white for the X axis, the second end on the second point. The distance
will then appear as shown below:

- Note that the measurements are NOT exported in your DWG file (see Step 4).

It’s a Snap! Take a photograph and create a 3D Model - Page 36

3.6 – Creating Lines and Polylines
You can create lines and polylines using 123D Catch Desktop Beta, in a way that is
very similar to the creation of Reference Points.

Creating a line or polyline in auto-match mode

To create a Reference Line or Polyline in auto-match mode, follow the steps below:
- Select your viewpoint: The preferred viewpoint is the one right in front of the
3D point that you want to create.
- Set your display mode.
o The mandatory visualization mode for creating Reference Points
is “Photo Lock”.
o You should then remove the 3D mesh from the display since the 3D
lines will be created with the pictures as references, not the mesh.
Uncheck the “Show Mesh” box in the sub-ribbon.

-
Click on in the tool bar and select “Create Reference Line”, or Edit >
Create Reference Line from the menu bar. A cursor then appears in the 3D
window.
- Click on the desired feature in the selected image to start your line or
polyline.

- The polyline will be completed by pressing “Enter,” “Escape,” or by a double

click.

It’s a Snap! Take a photograph and create a 3D Model - Page 37

 - The vertices of the lines or polylines are automatically computed in 3D using
the same “auto-matching” process than the one described for the creation
of 3D Reference Point (paragraph 3.7). This process may take some time,
depending on the number of vertices and the number or photos in your
project. During this process, the polyline is adjusted automatically:
o More points will be computed between two vertices by the auto-
match engine if the corresponding line follows a high contrast edge, to
make it more accurate.
o Starting from the 3rd line segment, the auto-match engine will check
if this line segment should not be in the same planar surface than the
one defined by the 2 first lines. If it is close to it, it will automatically
adjust this 3rd one to be co-planar with the first 2 ones.
- You may adjust these vertices the same way that you can adjust Reference
Points (see “Adjusting a Reference Point” in paragraph 3.7).
- The Reference Lines and polylines created with 123D Catch Desktop Beta
will be exported in the DWG file (see Step 4).

Creating a line or polyline in constrained mode

It may be useful to create lines or polylines by following the axis or planes defined
by your WCS:
- It will ensure the planarity of your set of lines, should they be planar.
- You can define planar sections.
- You can create 3D points which are not visible in the images, by locating
them at the crossing of 2 axes.

Example of defining a 3D point not visible in the image,

but located at the crossing of the X (red) and Z (blue) axis

- You can get a robust 3D point despite a weak confidence in the matching
process. Weak confidence may come from:
o The fact that the point is visible in 2 or 3 images only.
o The fact that the point is visible in images that offer very little
parallax.

It’s a Snap! Take a photograph and create a 3D Model - Page 38

The mandatory steps before using constraints when drawing your lines are:
- To set a reliable World Coordinate System,
- To well define your starting 3D point.

To create a Reference Line or Polyline in constrained mode, follow the steps below:
- Select your viewpoint: The preferred viewpoint is the one right in front of the
3D point that you want to create.
- Set your display mode.
o The mandatory visualization mode for creating Reference Points
is “Photo Lock”.
o You should then remove the 3D mesh from the display since the 3D
lines will be created with the pictures as references, not the mesh.
Uncheck the “Show Mesh” box in the sub-ribbon.

-
Click on in the tool bar and select “Create Reference Line”, or Edit >
Create Reference Line from the menu bar. Your cursor then appears as in the
3D window.
- Click on the desired feature in the selected image to start your line or
polyline.
- The following shortcuts are available to work in this mode:
o A: Toggle auto-snapping to axes
o Shift: Lock/unlock to axis constraint
o Tab: Cycle constraints
o Ctrl : To avoid any new snapping
o ESC: Exit the tool and validate the line created
o Backspace: Delete the previous segment of the line

- You can select which constraint to use by either right clicking in the 3D view
or pressing “Tab” several times.
o Right clicking will pop-up the following window:

It’s a Snap! Take a photograph and create a 3D Model - Page 39

 o Pressing Tab will allow you to loop between the constraints.

Example of constraint activated on X axis

Example of constraint activated on XOZ axis

It then becomes easy to draw lines and polylines along the edges of a building
for example, if your WCS has been properly set of course. In the following
example, drawing the contour of the windows on this façade was made possible
using this technique.

It’s a Snap! Take a photograph and create a 3D Model - Page 40

It’s a Snap! Take a photograph and create a 3D Model - Page 41
Step 4 – Export of the Photo Scene
You can export the entire Photo Scene or a selection of its components, in DWG, OBJ,
LAS, IPM or 3DP format.

- DWG is the abbreviation for DraWinG, the native drawing file format for
Autodesk applications. DWG files created by 123D Catch Desktop Beta can
be read in any Autodesk software starting from 2010 version. They contain 2
layers:
o Reference Points and Lines
o Reference Points Labels
Should you want to export the 3D Point Cloud as well, you will need to export it
separately in LAS format. See below.
- Autodesk® FBX® asset exchange technology facilitates higher-fidelity
data exchange between several Autodesk content creation packages. For
more information on the FBX format, visit the FBX SDK and Plug-ins page
at the Autodesk web site: www.autodesk.com/fbx. The FBX file created by
123D Catch Desktop Beta contains the 3D cameras, the 3D mesh, and all the
Reference Points, Lines and Labels created by the user.
- The OBJ is a 3D geometry definition file format with .obj file extension, which
contains the photo textured 3D mesh produced by 123D Catch Desktop Beta.
- The LAS file format is a public file format for the interchange of LIDAR data
between vendors and customers. This binary file format is an alternative to
proprietary systems or a generic ASCII file interchange system used by many
companies. More about ASPRS and LAS format on: http://www.asprs.org
The LAS file created by 123D Catch Desktop Beta contains the 3D point cloud
that is automatically extracted from the pixels of the Source Photos, during
the mesh creation process.
- IPM is the Inventor Publisher Mobile format from Autodesk®. The IPM
Viewer allows you to interactively view fixed or animated 3D assembly
instructions created with Autodesk Inventor Publisher software. You can
freely download this viewer for your iPhone, iPad, or iPod Touch (iOS 3.2 or
later), on:
http://itunes.apple.com/fr/app/inventor-publisher-mobile/id393147903?mt=8
- 3DP is the native file format to save and load Photo Scenes. 3DP files will
contain the whole or part of your Photo Scene, depending on your selection.
By default, the entire Photo Scene will be exported, which means:
o Cameras
o Reference Points
o 3D mesh
o Reference lines

It’s a Snap! Take a photograph and create a 3D Model - Page 42

 o Distance measures

Exporting a Photo Scene

- To export a Photo Scene, click on File -> Export Scene As… in the menu bar.
This will pop-up the export window, in which you can select your file name
and export format.

Importing the Automatic Point Cloud in AutoCAD using the LAS format
Please, follow the steps below to import your Automatic Point Cloud in AutoCAD:
- Load the corresponding drawing first (*.dwg)
- Index the .las file to create a .pcg file:

- Wait for the indexation to be completed, i.e. once you see the following
tooltip appear:

- Attach the .pcg file by either:

o Clicking the tooltip link
o Importing the .pcg file or
o Selection “Attach” from ribbon button.

Important: Let the insertion point to (0,0,0) and scale to default.

It’s a Snap! Take a photograph and create a 3D Model - Page 43

It’s a Snap! Take a photograph and create a 3D Model - Page 44
Step 5 – Creation of a Movie
One way to share your 3D model with others is to create a movie from it. This movie
will be created from a camera animation made of a series of “keyframes” manually
set where you want. In-between frames will be automatically interpolated by 123D
Catch Desktop Beta during the rendering phase.

Stage 1 – Creation of the animation

- Select your display mode:

o Background color: See Annex 1 – Preferences – Colors.
o “Wireframed” or “textured” mode: the rendering of your movie will
lately be performed in one of these modes exclusively. Note that you
can render a first movie in one mode, and render another movie with
the other one, keeping the exact same camera path for both.
- Click on “Animation Path” to start creating your keyframes.

- This will open the Animation Path tool.

- Selecting “Create default animation path” will automatically create one

keyframe for each of the stitched photos. The animation path defined this
way will then go through each one of your cameras.
- If you do not want to use this default animation path, you can start defining
your own keyframes.
- Select your viewpoint in the 3D window, and then click on “Create frame
from current view” in the Animation Path tool.

It’s a Snap! Take a photograph and create a 3D Model - Page 45

- You can also right click in the 3D window and select “Create New Keyframe”
from the menu.

- The camera path that you set is displayed in the 3D window.

- At any time, you can select one keyframe in the Animation Path tool or in
the 3D window, and either update the selected keyframe with the current
viewpoint, or delete it.

It’s a Snap! Take a photograph and create a 3D Model - Page 46

- A default time of 2 seconds is set between each keyframe. You can change
this by clicking on the corresponding box in the Animation Path tool.

- By clicking on the arrow in the Animation Path tool, you can play your
animation in the 3D window.

- By clicking on the “+” in the Animation Path tool, you can insert a new
keyframe after the one selected in the animation path.

It’s a Snap! Take a photograph and create a 3D Model - Page 47

Stage 2 – Rendering the movie

- Once you have created your animation path, you can render the
corresponding movie in various formats and store it, either on your
computer or directly on your You Tube account.

- Selecting “Export Video” will pop-up the following window:

- Select the video parameters that you want, either selecting one of the presets,
or defining your own set, including the codecs.

It’s a Snap! Take a photograph and create a 3D Model - Page 48

- Click on “Render” to generate your movie.
- You can directly render your video and publish it on your You Tube account.

- You must have your own account on You Tube to publish your videos.
Creating an account can be done on: http://www.youtube.com/
create_account?next=%2Findex
- Selecting “You Tube” will pop-up the following window:

- Once you have entered your account and password information, Photo Scene
Editor will automatically connect to You Tube to check the authentication. A
message “Authentication successful” will appear after this real time check.
- Set the various requested parameters, and click on “Publish”.
- Once the rendering is done, Photo Scene Editor will automatically export
the rendered movie to You Tube. You will need to connect to your You Tube

It’s a Snap! Take a photograph and create a 3D Model - Page 49

account to check when the video is ready for viewing on-line.

It’s a Snap! Take a photograph and create a 3D Model - Page 50