Digital 3D modelling of dinosaur footprints by photogrammetry and laser scanning techniques: integrated approach at the Coste dell’Anglone tracksite (Lower Jurassic, Southern Alps, Northern Italy)

1. What they did

Fossil footprints are usually recorded with drawings, photographs and casts. Drawings are simplified and depend on who makes them, photographs depend on the light, and casting a whole site with hundreds of tracks is not possible. The authors tested two newer ways to make three-dimensional (3D) digital models of footprints.

In March 2007, hundreds of dinosaur tracks, mostly in long trackways of up to 50 m, were found at Coste dell’Anglone in the Southern Alps of Northern Italy. The team worked on one trackway, CA2, made of 29 three-toed tracks pressed into what was once soft mud.

• In the field. They measured and drew the tracks of the trackway in the traditional way.
• Photogrammetry. Each footprint was photographed 5 to 6 times from different angles with a calibrated 14 megapixel digital camera. Software matched points between the photos to rebuild the shape of the print in 3D.
• Laser scanning. A laser scanner that measures shape by triangulation (a ShapeGrabber SG1002), with the generator it needed, was flown to the site by a helicopter of the Autonomous Province of Trento. Five scans at 0.3 mm resolution were joined into one model.
• Comparing. They described the best-preserved footprint, CA2 17, from the field drawing, the laser model and the photogrammetric model, and compared the measurements and the two digital models.

2. What they found

The trackway

• Foot length in the field ranged from 22.70 to 27.30 cm and foot width from 12.30 to 17.69 cm. The average stride was about 1.30 m.
• The toe pads gave a phalangeal formula of 3:4:5, pointing to a medium-sized theropod, although the angles between successive steps (97° to 178°) are unusual for a theropod.
• Formulas from the literature gave the animal a hip height of 1.07 m, a body length of 4.28 m and a body mass of about 130 kg. The stride to hip height ratio showed a walk, at an estimated 0.80 to 1.16 m per second depending on the formula used.

Footprint CA2 17

• Field drawing: the whole middle toe (digit III) was clear, but the tips of the two side toes were poorly preserved. No claw traces were seen.
• Laser model: three pad impressions on the middle toe were the deepest parts. The outer toe (digit IV) was pressed in more deeply (12 mm) than the inner toe (digit II, 6 mm), and the deepest point was the middle pad of the middle toe (16 mm). A steep wall ran along the outer edge of the outer toe.
• Photogrammetric model: showed details not seen in the field, including a faint claw trace on the middle toe and another on the outer toe, and a probable pad at the back of the outer toe.
• Measurements: field and model values were close but not identical. Foot length was 24.61 cm in the field and 24.53 cm on the models; foot width 17.31 cm and 16.30 cm; the angle between the inner and middle toes 14° and 8.5°; between the middle and outer toes 24° and 23.6°; total spread 38° and 32.1°.
• The two digital models agreed closely: in the area of the track they differed by less than 1 mm on average.

3. What they concluded

• Combining photogrammetry and laser scanning gives faster, more detailed and more accurate study of tracks, including depth in very shallow tracks, and lets tracks be studied in the laboratory from any angle without the problem of light direction.
• The trackmaker’s weight was probably carried mainly on the middle and outer toes, and the wall on the outer edge of the outer toe may show that toe moving down and outward, as suggested for other Late Triassic and Early Jurassic footprints. Such details can help match a track to its trackmaker.
• Outlines drawn from 3D models are more objective and informative than traditional drawings, but still depend, to a lesser degree, on the person interpreting them.
• 3D models could be used to study how the ground deformed, to combine the tracks of a trackway into an idealised track, to build shared virtual collections of track types, and even to print physical copies.
• Laser scanners are expensive, bulky and slow; photogrammetry is cheap, portable and quick. Both produce comparable models.
• Future studies should still include data from traditional methods.

4. How tracking was or was not used

The study is about recording and reading fossil footprints. The authors identified and measured the tracks of one trackway in the field, interpreted the toe pads, claw traces and depth of the prints to identify the kind of trackmaker, and used stride and foot length to estimate the animal’s size, gait and speed. No trailing in the living-animal sense was involved, but the trackway was followed and measured as a sequence of 29 steps.

5. Tracker or observer certification

Not reported. The field drawings and interpretations were made by the research team, and no individual’s skill is described or tested. The authors note that drawings, and to a lesser degree outlines from 3D models, depend on the interpreter.

6. Other technologies, skills or innovations

• Close-range digital photogrammetry with a calibrated Kodak DSC Pro SLR camera (14 megapixel, 50 mm lens) and a multi-image surface matcher.
• A ShapeGrabber SG1002 triangulation laser scanner, calibrated before each session; scans aligned with Polyworks IMAlign software.
• Models delivered as 3D images, drawn models and contour maps in common file formats (.wrl, .dxf, .ply) that can be opened in several programs.

7. Implications for conservation (from the paper)

• 3D models can keep tracks, and whole tracksites, permanently in their recorded state, even after weathering wears the real tracks away.
• Laser scanning can record tracks on surfaces too steep to study in person, such as near-vertical cliffs.
• Models can be shared among researchers without visiting the site or making casts, and offered to the public through virtual visits on the web or in museum exhibitions.