Analysis of classification accuracy for pre-filtered multichannel remote sensing data

Noise is one of the main factors degrading the quality of original multichannel remote sensing data and its presence influences classification efficiency, object detection, etc. Thus, pre-filtering is often used to remove noise and improve the solving of final tasks of multichannel remote sensing. Recent studies indicate that a classical model of additive noise is not adequate enough for images formed by modern multichannel sensors operating in visible and infrared bands. However, this fact is often ignored by researchers designing noise removal methods and algorithms. Because of this, we focus on the classification of multichannel remote sensing images in the case of signal-dependent noise present in component images. Three approaches to filtering of multichannel images for the considered noise model are analysed, all based on discrete cosine transform in blocks. The study is carried out not only in terms of conventional efficiency metrics used in filtering (MSE) but also in terms of multichannel data classification accuracy (probability of correct classification, confusion matrix). The proposed classification system combines the pre-processing stage where a DCT-based filter processes the blocks of the multichannel remote sensing image and the classification stage. Two modern classifiers are employed, radial basis function neural network and support vector machines. Simulations are carried out for three-channel image of Landsat TM sensor. Different cases of learning are considered: using noise-free samples of the test multichannel image, the noisy multichannel image and the pre-filtered one. It is shown that the use of the pre-filtered image for training produces better classification in comparison to the case of learning for the noisy image. It is demonstrated that the best results for both groups of quantitative criteria are provided if a proposed 3D discrete cosine transform filter equipped by variance stabilizing transform is applied. The classification results obtained for data pre-filtered in different ways are in agreement for both considered classifiers. Comparison of classifier performance is carried out as well. The radial basis neural network classifier is less sensitive to noise in original images, but after pre-filtering the performance of both classifiers is approximately the same.

US NOAA Fisheries and UK SAHFOS CPR surveys: comparison of methods and data

The US National Oceanic and Atmospheric Administration (NOAA) Fisheries Continuous Plankton Recorder (CPR) Survey has sampled four routes: Boston–Nova Scotia (1961–present), New York toward Bermuda (1976–present), Narragansett Bay–Mount Hope Bay–Rhode Island Sound (1998–present) and eastward of Chesapeake Bay (1974–1980). NOAA involvement began in 1974 when it assumed responsibility for the existing Boston–Nova Scotia route from what is now the UK's Sir Alister Hardy Foundation for Ocean Science (SAHFOS). Training, equipment and computer software were provided by SAHFOS to ensure continuity for this and standard protocols for any new routes. Data for the first 14 years of this route were provided to NOAA by SAHFOS. Comparison of collection methods; sample processing; and sample identification, staging and counting techniques revealed near-consistency between NOAA and SAHFOS. One departure involved phytoplankton counting standards. This has since been addressed and the data corrected. Within- and between-survey taxonomic and life-stage names and their consistency through time were, and continue to be, an issue. For this, a cross-reference table has been generated that contains the SAHFOS taxonomic code, NOAA taxonomic code, NOAA life-stage code, National Oceanographic Data Center (NODC) taxonomic code, Integrated Taxonomic Information System (ITIS) serial number and authority and consistent use/route. This table is available for review/use by other CPR surveys. Details of the NOAA and SAHFOS comparison and analytical techniques unique to NOAA are presented.