Search

Abstract--Underwater cameras increasingly are being used on remotely operated, autonomous, or towed vehicles to provide fishery-independent survey data in areas unsuitable for bottom trawls. To observe and quantify avoidance and [...]

Abstract--We describe the application of two types of stereo camera systems in fisheries research, including the design, calibration, analysis techniques, and precision of the data obtained with these systems. The [...]

Spatial management of vulnerable benthic ecosystem components such as deep-sea corals and sponges requires adequate maps of their distribution. These maps are often based on statistical models of survey data. The objective of this project was to validate the predictions of existing presence or absence and abundance models of deep-sea corals and sponges in the Aleutian Islands that were based on bottom trawl survey data. Model validation was conducted by comparing bottom trawl survey model predictions to the observations of an in situ camera survey conducted at randomly selected locations. The measures of goodness of fit (area-under-the-receiver-operator-curve, AUC) for the bottom trawl survey model predictions of camera survey observations ranged from 0.59 to 0.77 (for sponges and coral, respectively) and indicated that the bottom trawl survey models predicted the probability of presence for corals accurately across the Aleutian Islands. The bottom trawl survey models explained as little as 3% of the variability in Stylasteridae density and up to 17% of the variability in coral density. These results indicate that models of deep-sea coral distributions based on presence and absence data from bottom trawl surveys can be accurate and can provide useful information for spatial management of these vulnerable taxa. However, for some other taxa, such as sponges, care should be taken interpreting the results of bottom trawl survey models. An interesting finding of this study was that the residuals from the bottom trawl survey model-camera density relationships were negative in areas that remained open to fishing after 2005, possibly indicating an effect of continued bottom trawling on the abundance of corals in these areas. This study highlights the importance of validating models of species distribution using independent surveys, so that the results can be used with confidence to support decision-making processes. [ABSTRACT FROM AUTHOR]

Fishery surveys that call for the use of single or multiple underwater cameras have been an emerging technology as a nonextractive mean to estimate the abundance of fish stocks. Tracking live fish in an open aquatic environment posts challenges that are different from general pedestrian or vehicle tracking in surveillance applications. In many rough habitats, fish are monitored by cameras installed on moving platforms, where tracking is even more challenging due to inapplicability of background models. In this paper, a novel tracking algorithm based on the deformable multiple kernels is proposed to address these challenges. Inspired by the deformable part model technique, a set of kernels is defined to represent the holistic object and several parts that are arranged in a deformable configuration. Color histogram, texture histogram, and the histogram of oriented gradients (HOGs) are extracted and serve as object features. Kernel motion is efficiently estimated by the mean-shift algorithm on color and texture features to realize tracking. Furthermore, the HOG-feature deformation costs are adopted as soft constraints on kernel positions to maintain the part configuration. Experimental results on practical video set from underwater moving cameras show the reliable performance of the proposed method with much less computational cost comparing with state-of-the-art techniques. [ABSTRACT FROM AUTHOR]

Live fish recognition is one of the most crucial elements of fisheries survey applications where the vast amount of data is rapidly acquired. Different from general scenarios, challenges to underwater image recognition are posted by poor image quality, uncontrolled objects and environment, and difficulty in acquiring representative samples. In addition, most existing feature extraction techniques are hindered from automation due to involving human supervision. Toward this end, we propose an underwater fish recognition framework that consists of a fully unsupervised feature learning technique and an error-resilient classifier. Object parts are initialized based on saliency and relaxation labeling to match object parts correctly. A non-rigid part model is then learned based on fitness, separation, and discrimination criteria. For the classifier, an unsupervised clustering approach generates a binary class hierarchy, where each node is a classifier. To exploit information from ambiguous images, the notion of partial classification is introduced to assign coarse labels by optimizing the benefit of indecision made by the classifier. Experiments show that the proposed framework achieves high accuracy on both public and self-collected underwater fish images with high uncertainty and class imbalance.