13-2 Camtrawl: a Combined Camera-Trawl System for High Resolution Non-Lethal Sampling of Marine Environments
A combination trawl and camera system known as CamTrawl can augumnet traditional resource survey trawl capture sampling in many situations. CamTrawl is a non-lethal image-based alternative that provides increased spatial resolution and the potential to sample a greater diversity of animals compared to trawls. The codend, or capture bag at the end of a trawl net, in CamTrawl is replaced by a stereo-camera system to record passing fishes and invertebrates as the exit the trawl. The trawl forward of the camera system concentrates fish in the water column and restricts avoidance to the cameras. The system consists of two high resolution machine vision cameras connected to a computer for acquisition control and image storage. A series of high power strobes illuminate the image area. The cameras are calibrated to enable precise animal measurements from synchronized image pairs. By providing the location at which specific animals occur along the trawling path, CamTrawl data are especially useful in interpreting water column acoustic backscatter by enabling animals to be identified from multiple acoustic layers in both vertical and horizontal dimensions. Additionally, many small or fragile animals that often travel through the trawl and are not retained in the trawl catch may be identified with CamTrawl images, so information about these often critically important components of the ecosystem can be obtained without the use of other specialized non-trawl equipment. Fish captured by trawls rarely survive, and thus trawl survey methods may be inappropriate in some areas where fish stocks are severely depleted by overfishing or there is concern over habitat impacts. To aid in analysis, automated software processing routines have been developed to detect, track and measure targets in the image data. Tracking individual targets ensures accurate animal counts and provides more accurate size measurements as multiple measurements of the same individual are possible. Ongoing software developments are aimed at automating species recognition to provide fully automated processing and greatly reduce the time and personnel needed to extract data from the images. The Cam-trawl will not completely remove the need for physical sampling in situations where species identifications are particularly ambiguous, where biological specimens are required, such as age determination and diet analyses, or where water clarity is poor. With ongoing development, however, the Cam-trawl is poised to become a standard marine surveying tool to provide a more holistic view of the marine environment, and improve the management of our marine resources.