Danelle E. Cline

A review of techniques for the identification and measurement of fish in underwater stereo-video image sequences

Underwater stereo-video measurement systems are used widely for counting and measuring fish in aquaculture, fisheries and conservation management. To determine population counts, spatial or temporal frequencies, and age or weight distributions, snout to fork length measurements are captured from the video sequences, most commonly using a point and click process by a human operator. Current research aims to automate the measurement and counting task in order to improve the efficiency of the process and expand the use of stereo-video systems within marine science. A fully automated process will require the detection and identification of candidates for measurement, followed by the snout to fork length measurement, as well as the counting and tracking of fish. This paper presents a review of the techniques used for the detection, identification, measurement, counting and tracking of fish in underwater stereo-video image sequences, including consideration of the changing body shape. The review will analyse the most commonly used approaches, leading to an evaluation of the techniques most likely to be a general solution to the complete process of detection, identification, measurement, counting and tracking.

Detecting, Tracking and Classifying Animals in Underwater Video

For oceanographic research, remotely operated underwater vehicles (ROVs) and underwater observatories routinely record several hours of video material every day. Manual processing of such large amounts of video has become a major bottleneck for scientific research based on this data. We have developed an automated system that detects, tracks, and classifies objects that are of potential interest for human video annotators. By pre-selecting salient targets for track initiation using a selective attention algorithm, we reduce the complexity of multi-target tracking. Then, if an object is tracked for several frames, a visual event is created and passed to a Bayesian classifier utilizing a Gaussian mixture model to determine the object class of the detected event

An Automated Visual Event Detection System for Cabled Observatory Video

The permanent presence of underwater cameras on cabled observatories, such as the Victoria Experimental Network Under the Sea (VENUS) and Eye-In-The-Sea (EITS) on Monterey Accelerated Research System (MARS), will generate precious data that can move forward the boundaries of understanding the underwater world. However, sightings of underwater animal activities are rare, resulting in the recording of many hours of video with relatively few events of interest. Towards this end, an automated visual event detection system is in development at the Monterey Bay Aquarium Research Institute (MBARI) to address the problem of analyzing cabled observatory video. This paper describes the overall design of the development of a system to process video data and enable science users to analyze the results.

Laser Raman spectroscopy used to study the ocean at 3600‐m depth

Making geochemical measurements in the deep ocean is fundamentally difficult. For this reason, century-old technologies using water bottles and cores for sample recovery still provide the basic tools. With the development of research submersibles and remotely operated vehicles (ROVs), however, new opportunities for sophisticated sampling and analysis have arisen.

Learning-Based Algal Bloom Event Recognition for Oceanographic Decision Support System Using Remote Sensing Data

This paper describes the use of machine learning methods to build a decision support system for predicting the distribution of coastal ocean algal blooms based on remote sensing data in Monterey Bay. This system can help scientists obtain prior information in a large ocean region and formulate strategies for deploying robots in the coastal ocean for more detailed in situ exploration. The difficulty is that there are insufficient in situ data to create a direct statistical machine learning model with satellite data inputs. To solve this problem, we built a Random Forest model using MODIS and MERIS satellite data and applied a threshold filter to balance the training inputs and labels. To build this model, several features of remote sensing satellites were tested to obtain the most suitable features for the system. After building the model, we compared our random forest model with previous trials based on a Support Vector Machine (SVM) using satellite data from 221 days, and our approach performed significantly better. Finally, we used the latest in situ data from a September 2014 field experiment to validate our model.

ODSS: A decision support system for ocean exploration

We have designed, built, tested and fielded a decision support system which provides a platform for situational awareness, planning, observation, archiving and data analysis. While still in development, our inter-disciplinary team of computer scientists, engineers, biologists and oceanographers has made extensive use of our system in at-sea experiments since 2010. The novelty of our work lies in the targeted domain, its evolving functionalities that closely tracks how ocean scientists are seeing the evolution of their own work practice, and its actual use by engineers, scientists and marine operations personnel. We describe the architectural elements and lessons learned over the more than two years use of the system.

Towards automating underwater measurement of fish length: a comparison of semi-automatic and manual stereo–video measurements

Underwater stereo-video systems are widely used for counting and measuring fish in aquaculture, fisheries, and conservation management. Length measurements are generated from stereo-video recordings by a software operator using a mouse to locate the head and tail of a fish in synchronized pairs of images. This data can be used to compare spatial and temporal changes in the mean length and biomass or frequency distributions of populations of fishes. Since the early 1990s stereo-video has also been used for measuring the lengths of fish in aquaculture for quota and farm management. However, the costs of the equipment, software, the time, and salary costs involved in post processing imagery manually and the subsequent delays in the availability of length information inhibit the adoption of this technology. We present a semi-automatic method for capturing stereo-video measurements to estimate the lengths of fish. We compare the time taken to make measurements of the same fish measured manually from stereo-video imagery to that measured semi-automatically. Using imagery recorded during transfers of Southern Bluefin Tuna (SBT) from tow cages to grow out cages, we demonstrate that the semi-automatic algorithm developed can obtain fork length measurements with an error of less than 1% of the true length and with at least a sixfold reduction in operator time in comparison to manual measurements. Of the 22 138 SBT recorded we were able to measure 52.6% (11 647) manually and 11.8% (2614) semi-automatically. For seven of the eight cage transfers recorded there were no statistical differences in the mean length, weight, or length frequency between manual and semi-automatic measurements. When the data were pooled across the eight cage transfers, there was no statistical difference in mean length or weight between the stereo-video-based manual and semi-automated measurements. Hence, the presented semi-automatic system can be deployed to significantly reduce the cost involved in ad

Learning-based event response for marine robotics

Robotic vehicles have become a critical tool for studying the under-sampled coastal ocean. This has led to new paradigms in scientific discovery. The combination of agility, reactivity, and persistent presence makes autonomous robots ideal for targeted sampling of elusive, episodic events such as algal blooms. In order to achieve this goal, they need to be deployed at the right place and time. To that end, we have designed and will soon deploy a shore-based event recognition technology to continuously monitor remote sensing imagery for algal blooms as targets for robotic field experiments. A Support Vector Machine underlies a field-tested decision support system which scientists will consult prior to deploying robots in the coastal ocean. Our aim is to target oceanographic field experiments for evaluation and verification.

Detecting, Tracking and Classifying Animals in Underwater Observatory Video

For oceanographic research, remotely operated underwater vehicles (ROVs) and underwater observatories routinely record several hours of video material every day. Manual processing of such large amounts of video has become a major bottleneck for scientific research based on this data. We have developed an automated system that detects, tracks, and classifies objects that are of potential interest for human video annotators. By pre-selecting salient targets for track initiation using a selective attention algorithm, we reduce the complexity of multi-target tracking. Then, if an object is tracked for several frames, a visual event is created and passed to a Bayesian classifier utilizing a Gaussian mixture model to determine the object class of the detected event.

The environmental sample processor (ESP) software design: software for detection and quantification of microorganisms

The Environmental Sample Processor (ESP) instrument has been designed by the Monterey Bay Aquarium Research Institute (MBARI) for ocean sampling and monitoring. The ESP is an in situ sampling and processing device that enables near real-time detection of specific microorganisms through the application of molecular probes. The intended use of ESP is a 1 to 3 month deployment in 50 meters maximum depth for detection of harmful algal blooms. The authors present an overview of the software architecture deployed on the ESP instrument. The ESP software design is applied on two prototype instruments with similar mechanical design, but different control electronics. Presented in this work is the software architectural framework used that allows for controlled start up, shutdown, task and event handling in a concurrent software environment. They discuss how object-oriented design patterns such as the Adapter pattern are used to solve design problems and how testing improved reliability. A description and examples are given of the flexible ESP macro language that allows scientists to automate chemical processing steps. And finally, an algorithm for DNA probe array image registration and data extraction involving low-pass filtering, connected components, rotational translation, and component recognition and interpretation is presented.