Abstract:
Images of marine snow particles were collected using the Shadowed Image Particle Profiling Evaluation Recorder (SIPPER) camera imaging system abroad the multiple research cruises during and after the Deepwater Horizon oil spill. Images were analyzed from eight stations in the vicinity of the oil spill near the Deepwater Horizon platform during spring 2010, at 10 stations during August and September 2010, at 19 stations during May 2011 and 2012, and at 37 stations during August and September 2011, 2012, 2013, and 2014. SIPPER-captured particles were analyzed to determine their elongation ratios and fractal dimensions, two parameters that may impact particle sinking rates and impact predictions from numerical models of marine snow aggregates (Dissanayake et al. 2018).
Data Parameters and Units:
Station Location excel file: Deployment name, Date [M/DD/YYYY], Local time in [HH:MM, 0-2400], Local Time Out [HH:MM, 0-2400], Latitude N [decimal degrees], Longitude W [decimal degrees], Maximum sampling depth [m].
Marine snow data excel files: Excel Filename is Cruise_Station. Column A (Cast = downcast); Column B, Depth = Depth where camera imaged particles (m) (1m depth intervals); Column C, Volume Sampled = Volume of water (m3) imaged for marine snow particles; Column D, ESDbinStart = beginning Equivalent Spherical Diameter (ESD) particle size (mm); Column E, ESDbinEnd = end ESD particle size; Column F, TotalNumberOfParticlesAtDepthESD = Total Number of Particles in that specific size bin and 1 m depth interval; Column G, Elongation = mean elongation ratio of all particles in that specific size bin and depth; Column H, FractalDim = mean fractal dimension of all particles in that specific size bin and depth; Column I, Abundance = number of particles per m3.
The imagery was collected aboard multiple research cruises. The cruises information:
Dates Cruise name Ship Chief Scientist Location
05/05/10-05/17/10 SMP751001 R/V Weatherbird Jones NGOM
5/29/10-6/3/10 GU100200 RV Gordon Gunter N/A NGOM
08/06/10-08/16/10 WB0810 R/V Weatherbird Hollander NGOM
9/10/10-9/15/10 SD010 MV Specialty Diver Andrew Remsen NGOM
05/03/11-05/09/11 WB0511 R/V Weatherbird Sue Murasko NGOM
09/20/11-09/28/11 WB0911 R/V Weatherbird Rebecca Larson NGOM
05/07/12-05/15/12 WB0512 R/V Weatherbird Leslie Schwierzke-Wade NGOM
08/1/12 - 08/10/12 WB0812 R/V Weatherbird Leslie Schwierzke-Wade NGOM
8/5/2013 -08/11/13 WB0813 R/V Weatherbird Heather Broadbent NGOM
8/6/2014 -08/12/14 WB0814 R/V Weatherbird Heather Broadbent NGOM
Please note NGOM = northern Gulf of Mexico
Methods:
A. SIPPER Field Deployments: The SIPPER camera imaging system was developed by the University of South Florida, Center for Ocean Technology and described in Samson et al. (2001) and Remsen et al. (2004). The towed platform carried several environmental sensors (CTD, oxygen, chlorophyll fluorescence, transmissometer). SIPPER used a high-speed Dalsa Piranha-2 line-scan camera and a pseudo-collimated LED generated light sheet to image the shadows and outlines of resolvable particles that passed through a 100 cm2 field of view. The operational optical resolution of the system is ~65 um. SIPPER was towed at speeds between 2-3 knots in an oblique profile through the water column, spending approximately equal amounts of time at each meter of depth between the surface and up to about 300 m. At stations with a bottom depth shallower than 300 m, SIPPER was towed within approximately 5 m from the seafloor. Imaging and environmental data were stored internally on a Firewire hard drive and processed upon retrieval of the SIPPER instrument from a deployment using a customized software package called the Plankton Image Classification and Extraction Software (PICES). PICES was used to extract images of interest, classify them using user-specified training libraries and to manage the SIPPER images and environmental data collection. Currently, the PICES manages information for over 137 million SIPPER images.
B. SIPPER Data Analyses: SIPPER data, including concurrently collected environmental instrument data, were offloaded to a desktop PC after every deployment. Using the Plankton Instrument Classification and Extraction Software (PICES) (Kramer et al. 2011), images greater than or equal to 250 pixels in total area (~0.5 mm equivalent spherical diameter or ESD) were extracted from the raw SIPPER data, preliminarily classified and automatically entered into MYSQL based image database. A custom database management application called PICES Commander was used to manage and classify the resulting dataset. Classification of SIPPER images involved several steps including both automated classification and manual labeling of the images.
(1) Image Extraction and Preliminary Classification: During the image extraction step, features for each image are determined and entered into a feature table in the PICES database and an initial classification for each image is predicted using a comprehensive, multi-class feature selection (MFS) support vector machine (SVM) (Kramer et al., 2011). This classifier was initially developed from training images collected in the Gulf of Mexico prior to the Deepwater Horizon oil spill. Training images and new image classes were added throughout the sampling period and slowly replaced the images that were collected before the spill. We utilized a hierarchical naming format for the particle and plankton image classes to reflect the level of similarity between image classes. More closely related image classes shared one or more parts of their name separated by underscores and this naming schema is used in the classification effort.
(2) Manual Validation: After image extraction and the first run of the comprehensive classifier, thumbnail images of the resulting classifications were then examined using the image browsing and validation capability of PICES Commander. Image classes that consisted of numerous true positive examples (for example calanoid copepods or detritus) were noted for inclusion in the final classification effort and then ignored while image classes that were less abundant or rare were closely examined. True positive image examples from these rarer image classes were validated, meaning that their image class label could not be relabeled during a subsequent classification. This was done for every predicted image class until all rare images were validated from the first comprehensive classification. Occasionally, a new image class was encountered. When this occurred, that class was added to the comprehensive classifier and any encountered examples would be validated and added to the training library.
(3) Run Final Comprehensive MFS-SVM Classification: After all deployments had their images extracted, initially classified, and new image classes encountered, labeled, and added to the training library, the final comprehensive MFS-SVM classifier was rebuilt and rerun on all deployments using all the new training examples that had been uncovered in Step 2. The purpose of this classification cycle was to identify which common classes to include in the ultimate dual classifier, identify deployments where less common image classes might be abundant enough to include in the ultimate dual classifier, and validate any new rare image examples that may have been identified during the second run of the comprehensive classifier.
(4) Build Deployment-Specific MFS-SVM Classifiers: For each deployment, after rare and uncommon groups were searched for true positives and validated in Step 3, an MFS_SVM classifier was pared down to the most common classes found in that deployment and rebuilt. New MFS classifiers were, therefore, built for each deployment with classes specific to that deployment. Multiple deployments could share the same final MFS classifier if they shared the same common groups. A final feature selection was then run for all the image classes in the final MFS Classifiers, using Amazon High-Performance Computing, to create the most appropriate feature set for the classifiers.
(5) Build Deployment-Specific BFS-SVM Classifiers: Another set of SVM classifiers was also built using binary feature selection (BFS) as described in Kramer et al. (2011), where features and parameters are tuned for each class pair combination. These BFS classifiers used the same image classes as the MFS classifiers so that they were specific for the particular conditions encountered during each deployment. By using pair-wise feature selection, overall classification accuracy can be improved, and feature selection and SVM training time can be reduced.
(6) Run Dual Classification: Lastly, these two sets of classifiers (MFS-SVM and BFS-SVM) were used to run a modified version of the dual classifier described for the video plankton recorder (VPR) by Hu and Davis (2006). As the name implies, a dual classifier makes use of two classifiers where each classifier makes its own separate prediction and a particle image is only labeled as a specific image class if both classifiers agree, otherwise the particle image is labeled “other”. In the case of the VPR, a SVM classifier using texture-based features and a neural net using shaped-based features were used as the two classifiers. For the SIPPER data, a neural net performed poorly and it was determined that a SVM BFS using separate feature sets for each pairwise classification was the best alternative. Along with the SVM MFS classifier, the SVM BFS made up the SIPPER dual classifier. A further modification took advantage of the hierarchical naming structure of SIPPER image classes by allowing for predictions from partial matches between the two classifiers. For example, if one classifier predicted that an unknown image was crustacean_copepod_calanoid and the other classifier predicted it was crustacean_copepod_oithona, it was very likely that the image was, in fact, a copepod. Rather than labeling this disagreement as “other” as the traditional dual classifier would have done, our modified classifier would label the image as crustacean_copepod since both classifiers agreed to that partial match. Only when the two classifiers predictions shared no common root in the SIPPER class names (e.g. one classifier predicts an unknown image is detritus_snow and the other predicts larvacean) would an image be labeled “other”. Dual classifiers are used because they improve specificity (reduce the rate of false positives) and thereby give more accurate abundance estimations, especially in regions of low relative abundance. Dual classifiers were run on each deployment using the most common image classes in that deployment as determined in Step 3. A total of 32 individual image classes were common enough to be used in at least one deployment and are listed below. Validated images from rare and uncommon classes that were removed from the classifier retained their validated labels in the classifier output.
C. Particle Characteristics: Two important parameters in coagulation models are particle shape and texture (Dissanayake et al. 2018). The shape of aggregate affects its settling velocity and the fractal property of a particle affects its interactions with other particles. For a certain mass of an aggregate, its porosity, density, settling velocity, and collision frequency with other particles are affected by its fractal dimension (Burd and Jackson 2009). The methods are reported in Dave (2018) and briefly below.
Elongation ratio: We used elongation to measure the shape of a particle’s closeness to a circle. The ISO 9276-6 suggests that the aspect ratio should not be used for needle-like structures (Olson 2011). Since a number of images were long and thin, we used the recommended formula for elongation as shown below:
Elongation ratio = 1 – (minor axis/major axis)
Fractal Dimension: Fractal dimension was used to measure the texture of an aggregate. We used the box-counting method (Kilps et al. 1994) with suggested modifications by Barton and La Pointa (1995), who recommended using a box with the same aspect ratio as that of the object instead of a square box. That means dividing the whole object into nxn boxes and recording the number of intersecting boxes N(n). The fractal dimension is then estimated by the slope of the least mean square linear regression of the log-log plot of N(n) vs n. The implementation of the box-counting method showed that the error range using standard mathematical fractal images, like the Sierpinski triangle, Koch curve, and the Dragon curve, were within an acceptable range of -0.079 to 0.053 (Dave, 2018).
Provenance and Historical References:
Barton, C. C., Paul, R., & Pointe, L. (Eds.). (1995). Fractals in the earth sciences (pp. 141-178). New York: Plenum press.
Burd, A. B., & Jackson, G. A. (2009). Particle aggregation. Annual review of marine science, 1, 65-90.
Dave, P. P. (2018). A Quantitative Analysis of Shape Characteristics of Marine Snow Particles with Interactive Visualization: Validation of Assumptions in Coagulation Models. Graduate Theses and Dissertations. https://scholarcommons.usf.edu/etd/7279
Dissanayake, A. L., Burd, A., Daly, K. L., Francis, S., & Passow, U. (2018). Numerical Modeling of the Interactions of Oil, Marine Snow, and Riverine Sediments in the Ocean. Journal of Geophysical Research: Oceans. doi:10.1029/2018jc013790
Hu, Q., & Davis, C. (2006). Accurate automatic quantification of taxa-specific plankton abundance using dual classification with correction. Marine Ecology Progress Series, 306, 51–61. doi:10.3354/meps306051
Kilps, J. R., Logan, B. E., & Alldredge, A. L. (1994). Fractal dimensions of marine snow determined from image analysis of in situ photographs. Deep Sea Research Part I: Oceanographic Research Papers, 41(8), 1159–1169. doi:10.1016/0967-0637(94)90038-8
Kramer, K., Goldgof, D.B., Hall, L.O., A. Remsen (2011). Increased classification accuracy and speedup through pair-wise feature selection for support vector machines. IEEE Symposium on Computational Intelligence and Data Mining (CIDM), Paris, 2011, pp. 318-324. doi: 10.1109/CIDM.2011.5949457
Olson, E. (2011). Particle shape factors and their use in image analysis part 1: theory. Journal of GXP Compliance, 15(3), 85.
Remsen, A., Hopkins, T.L., S. Samson (2004). What you see is not what you catch: a comparison of concurrently collected net, Optical Plankton Counter, and Shadowed Image Particle Profiling Evaluation Recorder data from the northeast Gulf of Mexico. Deep Sea Research Part I: Oceanographic Research Papers, 51(1): 129-151. doi: 10.1016/j.dsr.2003.09.008
Samson, S., Hopkins, T., Remsen, A., Langebrake, L., Sutton, T., and J. Patten (2001). A System for High-Resolution Zooplankton Imaging. IEEE Journal of Oceanic Engineering, 26(4): 671-676. doi: 10.1109/48.972110
View Metadata