Import point cloud - PIX4Dmatic

PIX4Dmatic supports the import of external point clouds, regardless of the sensor used to capture them or the software used to process them. This article provides step-by-step instructions for importing a point cloud into PIX4Dmatic.

 

What is a point cloud?

A point cloud is a dense set of data points in three-dimensional (3D) space used to represent the external surfaces of terrain, structures, or objects. Each point typically contains X, Y, and Z coordinates, and may also include additional attributes such as color, reflectance intensity, or classification information.

How is a point cloud generated?

Point clouds are typically generated using one of the following methods:

• Photogrammetry: Uses overlapping photos taken from different angles (nadir, oblique) to reconstruct a 3D model and extract a point cloud.

• Terrestrial laser scanning (TLS): Static measurement system based on laser from a fixed position. 

• LiDAR (Light Detection and Ranging): Similar to TLS but mounted on a drone aircraft, or vehicle. It is widely used for mapping large areas.

What are the normals in a point cloud? 

Each point in a point cloud represents a precise location in three-dimensional (3D) space. However, it does not inherently convey information about the orientation of the surface at that point. This is where the concept of normals becomes essential.

A normal is a unit vector that is perpendicular to a surface at a specific point. In 3D modeling and point cloud processing, normals are fundamental for understanding the local geometry and orientation of surfaces. They are widely used in tasks such as surface reconstruction, registration, shading, segmentation, and object recognition.

Since point clouds consist only of discrete points without any explicit surface connectivity (unlike meshes or CAD models), normals are not directly available. Instead, they must be estimated by analyzing the spatial distribution of neighboring points.

How can I import a point cloud?

1. Importing a point cloud in PIX4Dmatic can be accomplished with either of the two methods below:

   1. Drag and drop onto the Home screen: Simply drag and drop the point cloud onto the Home screen, and PIX4Dmatic will automatically generate a new project using the imported point cloud.
      
   2. File > Import > Point clouds...: For an existing project, select this option to brows and import the desired point cloud.
      
      
      

2. Select Acquisition type and Coordinate Reference System.

After the point cloud is imported, PIX4Dmatic will ask for the Acquision type and the coordinate reference system (CRS) of the point cloud. Establishing the CRS of the point cloud is necessary in all acquisition types. If the CRS information is on the header of the point clouds, the CRS is pre-filled with this information.

The Acquisition type is determined by the method used to capture the point cloud, and there are four possible Acquisiton types when importing a point cloud:

• Unstructured: Normals cannot be computed for point clouds with this acquisition type. Mesh generation will not be possible.
• Nadir: Normals will be calculated assuming the laser scan was captured from above.
• Static: The scanner is assumed to be positioned at the origin of the point cloud's local coordinate system. The origin of the local coordinate system is defined in the offset field of the LAS header.
• With trajectory: A trajectory file can be imported together with the point cloud. To use the trajectory option, the point cloud must include a GPS time attribute. If "With trajectory" is selected, follow the steps below:

How to import a point cloud with a trajectory file?

The trajectory file of a LiDAR flight (usually from an airborne system like a drone or aircraft) contains detailed information about the path followed by the sensor during data acquisition. It is essential for georeferencing the LiDAR data. 

These trajectory files typically include the following information:

• GPS time. Indicates the exact moment of each recorded position.
• Position. X, Y, and Z coordinates of each timestamp in a global or local coordinate system.
• Orientation. Describes the orientation of the sensor on each measurement (roll, pitch, and yaw). 
• Velocity. Velocity of the airborne system in m/s.
• Accuracy. PDOP, number of satellites, etc.

Currently, PIX4Dmatic imports data related to GPS time and positional information. After selecting and importing the point cloud and choosing the acquisition type as "With trajectory," the next step is to configure the parameters for importing the trajectory file.

PIX4Dmatic supports trajectory files in the .csv format.

• [1] Trajectory file: Select the trajectory file.
• [2] Trajectory coordinate system.
  • Point cloud CRS. 
  • Point cloud local. 

• [3] Column separator: Defines the column separator as a comma, semicolon, pipe, space, or tab.
• [4] Decimal separator: Defines the decimal separator as a comma or period. 
• [5] Skipped rows: Specifies the number of rows to be skipped if the trajectory file includes a header.
• [6] Column format:
  • TXYZ: GPS time, X, Y, and Z.
  • XYZT: X, Y, Z, and GPS time.
• [7] Coordinate reference system (CRS). Specify the CRS to be used when importing the point cloud. If the acquisition type is set to With trajectory, both the point cloud and the trajectory files must be in the same CRS.

Upon successful importation of the point cloud and trajectory file, PIX4Dmatic will display both the point cloud and the trajectory within the project interface.

How can I visualize the point cloud?

The imported point cloud can be colorized using different visualizations:

• By default, colorization is based on RGB if the dataset was captured and colorized using RGB images.

  
• Color by elevation. Colorization based on the altitude of each point of the point cloud.
  
• Color by normal. Colorization based on the X, Y, Z components of the point normals, mapped to R, G, B, respectively. Normals are computed at import based on the acquisition type (except for Unstructured).
  
• Color by GPS time. Applies a rainbow gradient to the point cloud based on the GPS time. Points recorded earliest are colored red, transitioning through the spectrum to purple for the latest recorded points.