3D Capture Technology - Passive Stereo Photogrammetry
-
Passive Stereo Photogrammetry
Dimensional Imaging has pioneered the 3D surface image capture method of Passive Stereo Photogrammetry.
This allows extremely dense 3D models to be recovered simply from a stereo pair of standard digital still camera images. It does not require any pattern projection (structured light or active stereo photogrammetry) onto the scene or any form of laser scanning.
As a result the capture process is extremely fast and as simple as taking an ordinary photograph: one or more stereo pairs of images of a subject or scene are captured simultaneously and instantaneously using an array of pre-calibrated digital stills cameras. -
Dense Range Maps
DI3Dcapture™ automatically processes stereo pairs of images (using camera calibration data) to create dense and accurate range maps.
A range map is a highly efficient image-like structure for storing dense 3D data. Every pixel in the range map image represents the 3D position of that pixel in space.
DI3Dcapture™ is therefore capable of calculating the 3D position of every pixel in the input images: up to 10 million 3D points per stereo pair or more depending on the resolution of the cameras used. -
Automatic Merging
It is often not possible to capture all of the required surface from a single point of view. DI3Dcapture™ offers a simple solution: use multiple stereo pairs of cameras to capture simultaneously 3D surface patches from multiple view-points.
DI3Dcapture™ then allows the multiple surface patches to be merged automatically together to form a single seamless 3D mesh.
-
Variable Density Polygonisation
In order to be displayed using current 3D graphics engines and opened using most third party 3D software, DI3Dcapture™ converts range maps into a polygonal meshes in most standard 3D mesh file formats.
The user can control the density of mesh in order to optimize it for speed of display and manipulation (low density) or high resolution (high density).
In addition, the high resolution images can be texture mapped onto the polygon mesh, to provide a full color, high resolution, highly realistic 3D image surface.
DI3Dview™ has been specially developed to allow the highest resolution DI3D™ 3D surface images to be viewed at full definition so that every detail is clearly displayed. -
Advantages of Passive Stereo Photogrammetry Over 3D Scanning
The technique of 'scanning' is usually associated with lasers and 'laser scanning'. Scanning involves passing some sort of sensor (in most cases a laser) over the object to acquire the object's shape.
3D scanning (particularly 3D laser scanning) is recognised as an extremely accurate form of 3D surface acquisition and is mainly used for acquiring the 3D scans of inanimate objects.
However, the 3D scanning process is not instant, and so the process of scanning is not always suitable for some situations such as capturing the 3D scan of living organisms, as it is usually impossible for the subject to stay completely still during the scanning process, causing inaccuracies in the final 3D surface image or 3D scan.
When scanning living objects such as the human face, head or body, Passive Stereo Photogrammetry has several advantages over 3D scanning:
1) Instant Capture: The acquisition phase of the Passive Stereo Photogrammetry process is instant and so it is possible to capture lving organisms and moving objects.
2) High Resolution Images: The input to the Passive Stereo Photogrammetry process is high resolution images, The relationship between every 3D point and the high reolution images is known so it's possible to capture detialed texture image information at the same time as the 3D image or 3D scan information. -
Advantages of Passive Stereo Photogrammetry Over Active Stereo Photogrammetry
Active Stereo Photogrammetry is similar to Passive Stereo Photogrammtery, however Active Stereo Photogrammetry relies on the use of Structured Light to acquire the 3D surface image or 3D scan.
Using structured light involves projecting a pattern onto the subject (either a random speckled pattern or a series of stripes) and that pattern is used to help obtain the 3D scan.
Active Stereo Photogrammetry often has problems scanning subjects with darker skin tones, hair and facial hair because the the pattern cannot be seen in these areas, and so no 3D surface can be calculated for them, resulting in large holes in the final 3D scan
When using structured light, in order to obtain proper texture images, the pattern must somehow be removed, this can be done wither by processing the images to remove the pattern before they are applied to the 3D scan (which usually results in a decrease in quality of the texture images) or the texture images are captured at a slightly different time from when the 3D scan was obtained (resulting in a slight inconsistency between the geometry of the 3D scan and the 2D texture image).
Passive Stereo Photogrammetry has several advantages over Active Stereo Photogrammetry:
1) Quality Texture Images: The acquisition phase of the Passive Stereo Photogrammetry process does not require structured light, and so there is a true one to one correspondence with every colour pixel and every 3D point, the original texture images are always of the highest quuality.
2) No Holes: Again, because the acquisition phase of the Passive Stereo Photogrammetry process does not require structured light, there are no holes in the final 3D scans in area such as the hair.
