科研之友
微信
新浪微博
Facebook
分享链接
摘要
In order to understand physiological, ecological and biological processes, it is often crucial to determine an organism's volume and surface area (SA). Most of the available methods require sacrificing the organism or at least removing it from its natural habitat, in order to measure these parameters. Advances in computer vision algorithms now allow us to determine these parameters using non-destructive, three-dimensional modelling. The addition of cloud computing and the availability of freeware make this tool widely accessible. Photographs of corals and sponges were taken in natura and used to create digital 3D models using the structure-from-motion' technique. Modelling was done online using 123D Catch freeware (Autodesk Inc.). Volume and SA of the corals and sponges were calculated from these 3D models. Comparing in situ 3D modelling to current measuring methods (e.g. water displacement, paraffin dipping) showed that volume calculation by 3D modelling gave fast results accurate to within 8% of estimated true volume. Using cloud computing enabled the creation of a 3D model in <30min. SA accuracy was found to differ significantly, depending on the shape of the modelled object, with an accuracy ranging widely from 2% to 18%. We found that in situ volume and SA measurements created by 3D modelling enable easy, fast and non-intrusive studies of benthic aquatic organisms, without removing the subject organisms from their habitat, thus enabling continuous study of natural growth over extended time periods. The freely available freeware, along with ease of use, makes this method accessible to many areas of research.
- 出版日期2015-5
