High throughput photogrammetric measurement of morphological traits in free-ranging phototropic insects
Mungee, M.; Athreya, R.
Show abstract
O_LIRemote measurement of morphological traits in free-ranging animals is very useful for many studies, but such non-invasive photogrammetric methods are limited to large mammals and have yet to be successfully applied to insects which dominate terrestrial ecosystem diversity and dynamics. Currently, insect traits are measured using collected specimens; the process of collection and maintenance of specimens can impose a heavy and unnecessary cost when such specimens themselves are not needed for the study.\nC_LIO_LIWe propose a rapid, simple, accurate, and semi-automated method for high-throughput morphometric measurements of phototropic insects from shape and size calibrated digital images without having to collect a specimen. The method only requires inexpensive, off-the-shelf, consumer equipment and freely available programming (R) and image processing (ImageMagick) tools.\nC_LIO_LIWe demonstrate the efficacy of the method using a data set of 3675 images of free-ranging hawkmoths (Lepidoptera: Sphingidae) attracted to a light screen. Comparison of trait values from a subset of these images with direct measurements of specimens using a scale showed a high degree of correspondence. We have also identified several error metrics which help in assessing the method in an objective manner.\nC_LIO_LIAlthough this method was developed for nocturnal phototropic insects, it can be used for any other (small) animal that can be imaged on a simple graph paper. While this technique will be generally useful for a variety of studies of insect traits, we suggest that it is particularly suited as a commensal on multi-epoch and multi-location population monitoring of insects in the context of climate and land-use change, where repeated sampling obviates the necessity of collecting specimen every time. It will help in accumulating a large amount of reliable trait data on hundreds of thousands of individual insects without an overwhelming expenditure on collection, handling, and maintenance of specimens.\nC_LI
Matching journals
The top 2 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
Similar papers in this journal
Similar papers in this journal
- Insect Size Matters: Using Image and Dimensions Together Improves Image Classification 94%
- Exploiting facial side similarities to improve AI-driven sea turtle photo-identification systems 94%
- Comparison of Two Individual Identification Algorithms for Snow Leopards (Panthera uncia) after Automated Detection 94%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.