Alternate Title

Accurate 3-D Morphological Measurement Using a Structured-Light Range Sensor

Document Type



A single-plane structured light range-sensor was tested to establish its usefulness in acquiring 3-D measurements of fish skulls. Twenty-one distances among 22 landmark points for each of 12 neurocrania of the scorpaenid fish Neomerinthe hemingway were taken with digital calipers, with a video-based 2-D imaging system widely used in systematic studies, and with a single-plane structured-light range sensor of inexpensive and simple design. Measures taken by 3-D sensor are highly correlated with those obtained from caliper measurement (r = 0.9995, P << 0.001 ), with a precision ranging from 0.08–0.43 mm. Like caliper-based measurements, they are less strongly correlated with measurements derived from projected video-imaging. Most skulls were scanned in just over 3 minutes each. Range maps, typically establishing the (x,y,z) coordinates of more than 75,000 points per scan, can be obtained in about 40–50 CPU seconds using software running on multiple platforms. Sensor data taken from different views can be merged to build a more complete 3-D reconstruction. System design, calibration, and use are discussed. By eliminating error due to perspective effects inherent in measuring from projected video images, such sensors hold considerable promise in quantifying biological shape in 3-D for comparative and functional studies.

First Page


Last Page