The images on this page compare conventional and fluorescence photographs of experimental surfaces being used in a benthic ecology experiment. The white-light photographs were made using a Nikonos V fitted with a 28 mm lens and the Nikonos close-up kit, with a Nikon SB104 strobe. The fluorescence photographs were made with a Nikonos V fitted with the same lens set-up, but mounted in a custom 'DarkBox' that excludes ambient light and enables fluorescence photography in daylight. The DarkBox holds two Ikelite Substrobe 50's fitted with NightSea BE1 blue-light excitation filters. The camera was fitted with a NightSea BB59 barrier filter to block the reflected blue light.

White-light photograph ©Robert Miller	Fluorescence photograph ©Charles Mazel

Image Classification

We applied computer classification rules, using the red and green signal levels, to pick out features of interest in the fluorescence image above. The results of the image classification are shown in the figure below. The scale bar in the upper right corresponds to approximately 1 cm.

Classes in image to right red - chlorophyll-only fluorescence from green or brown algae yellow - chlorophyll plus phycoerythrin fluorescence from red algae green - bright green fluorescing features. Some are from strong reflections from bright objects (the rejection of excitation light is not quite 100%), but for others the source is unknown. white - bright yellow fluorescing features, presumably phycoerythrin in red algae in which the photosynthetic pathway is breaking down

Once pixels have been assigned to classes it is straightforward to compute the area of the surface covered by each type, and to investigate spatial distributions. It is also possible to distinguish filamentous from leafier-type algae in the fluorescence imagery, and with some local knowledge one could possibly assign the individual algae to genus or species.

Detailed sub-images

The following three image triplets show, left to right, details of portions of the white-light image, the fluorescence image, and the assignment of image features to classes.

The experiment

Robert Miller, a graduate student at University of Massachusetts, Boston, deployed a set of granite settlement plates at a depth of approximately 10 meters at Shags Rock, Nahant, Massachusetts, just offshore of the Northeastern University Marine Science Center. The surfaces were exposed to a range of factors (shading, protection from grazers, etc.) to study invertebrate settlement processes.