Check it out! – Click, Drag and Zoom the 3D model.

In the interactive window below is a simple example of 3D point cloud model. In the first example the settings can change the model from a point cloud to a mesh model. Additionally, the size of each point can be adjusted as well as how many millions of points at one time are being cached.

Why is the point cloud so pixelated?

Initial impressions of a point cloud are usually fairly anti-climatic. We are often asked why the model its so pixelated. Once we explain the technology and what a 3D point cloud is, then eyes get big and there is a genuine expression of understanding. Creating a “digital 3D twin” is a miracle of math that is impressive to say the least.

Let us explain, what you’re seeing in the 3D point cloud above are millions of textured pixels that were derived from high definition pictures taken from a drone that flew a preplanned path over the project and gathered hundreds, perhaps even thousands of images of the area below.

As each image was taken the drone encoded the image with metadata that included the GPS location it was taken from, the angle of the gimbal, the focal length of the camera and other key information about the exposure.

Once the drone flight was finished, the images are put through post processing. Back in the office all the pictures are uploaded to a project where our software can analyze and calculate the location of each point of the 3D point cloud.

In a grossly over simplified explanation, the software “mines” each image’s metadata for its GPS coordinates, the angle of the image projection and other information that was stamped at the time it was taken. We also add known control points to the project which are GPS coordinates of official targets that are included as a part of the flight plan.

From there the software goes to work triangulating the location of each point of the cloud in 3D space by finding like pixels of color and metadata (i.e. GPS coordinates, gimbal angle, lens aperture. etc.) from adjacent images and matching that with as many pixels from the specific image its processing.

Calculating and analyzing all of this data, the software derives the exact georeferenced location of as many pixels as possible in real space from each image. With each high resolution image of 20 million pixels each can you imagine the math that needs to take place to locate each point?

Combined all together, these pixels/points create a “cloud”, a “3D point cloud” or “digital twin” of the project. With each known point representing a particular coodinate in space, the more points, the more dense the point cloud will become and the more it can emulate smooth and less pixilated geometry.

Ultimately all of this effort can produce final outputs that can be exported in a number of ways including the 3D point cloud itself, a 3D mesh, digital elevation models (DEM), topographic line geometry, orthomosaic GeoTIFFS and many other engineering files.