UC San Diego

Ecologists are interested in studying fish length distributions as a metric for the health of fish populations, both for fishery management and to gauge the effect of policies and worsening ocean conditions. Current methods of gatheringthese data are time intensive, and require expensive equipment and/or training. Approaching this issue with citizen science -- utilizing efforts from the general public -- allows us to gather much more data about fish populations than what can currently be acquired, at a reasonable cost. We are working towards developing a device and software platform that allows recreational divers to take pictures of fish, and have the images processed such that length and species information are determined automatically.

This work focuses on the mechanism through which fish length is calculated. The device incorporates a standard laser pointer with a consumer waterproof camera, and uses the location of the laser dot in the object to determine the distance between the object and the camera. We also discuss methods to ensure that the parameters of the laser with respect to the camera are known, as well as how we can ensure that these procedures can be performed in the field. We test these by conducting trial calibrations and length measurements in a controlled setting, and evaluating where sources of error come from. From our testing we are able to verify that our technology is capable of obtaining measurements within 20% of of true length, similar to the error margins of visual estimation.