APL-UW researchers developed an automated system to determine fish size quickly, in situ, and without human intervention, using an off-the-shelf video camera and custom measurement algorithms on a control computer. Mounted on a fishing vessel's discard chute, the system measures the size of every halibut that is inadvertently caught prior to its return to the sea. The system speeds the required measurements, lessens the time halibut are out of the water, and thereby reduces mortality.

The APL-UW system was deployed aboard a factory trawl vessel in the Bering Sea for two months during winter 2017. With no prior testing on live halibut, the system was able to measure over 2000 fish in an actual operational environment. The hardware and computer networking design were robust, handling the extreme conditions of winter in the Bering Sea. With this successful first at-sea test, there is great promise for the autonmated fish measurement system and its adoption by the fishing fleet.

Problem

Halibut are caught accidentally by factory trawlers. By law they must be released back to the ocean. The goal is to keep the halibut alive, but because the law also requires that at least every 5th fish must be measured before release, they are on deck longer, which increases mortality. Can the science (accurate measurements) and conservation (rapid release) objectives both be met?

Solution

Multi-frame video is an attractive, cost-effective method to address electronic monitoring and measurement of fish. Measurement by video camera is both accurate and rapid, fulfilling both the scientific and conservation goals. It is more robust than a still image solution because of the sheer number of images captured by the video camera operating at 30 frames per second as the fish traverses the camera's field of view.

The system was installed aboard a U.S. Seafoods factory trawler. It was placed directly in the exit slide of the discard chute, which puts it immediately downstream the measurement table used by observers to perform manual fish measurements. Operation was very simple: crew memebers simply flip a power switch to begin and turn it off when the haul's worth of discard is complete.

How It Works

The measurement algorithm has three individual components: segmentation, tracking, and measurement.

Segmentation is the process by which the image of the fish is distinguished from the background. In the tracking step, a Kalman filtering tracker estimates the position and velocity of the fish and associates the segmentations of the same fish in contiguous frames. These segmentations are then rotated into the horizontal based on their orientation. In the measurement step, a scoring function is used to locate the head and tail endpoints of the final segmentation using scoring functions. The length of the fish is then calculated by connecting the head and tail endpoints through a morphological midline along the center of the fish image body.

Cultivating Student Researchers

Much of the development and testing was done by undergraduate students at the University of Washington. They represent a wide cross-section of disciplines including oceanography, electrical engineering, mechanical engineering, and computer science. The students presented the system and its purpose to peers and faculty. They demonstrated the fish measurement system and discussed their roles on the project to several university and high school student groups.

Partnerships

APL-UW researchers partnered with U.S. Seafoods, a Seattle-based company whose vessels particpate in Amendment 80 fisheries, the management program targeted for the deck-sorting Exempted Fishing Permit (EFP). U.S. Seafoods provided access to their dockside facilities, loaned APL-UW fish slide equipment from their inventory, and ultimately hosted the measurement system itself aboard their factory trawl vessel, Seafreeze America.