Textural analysis and structure-tracking for geological mapping: applications to sonar images from Endeavour Segment, Juan de Fuca Ridge

Blondel, Philippe; Sempéré, Jean‐Christophe; Robigou, V.

doi:10.1109/oceans.1993.326187

Cited by 13 publications

(11 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This shows that the textural entropy and homogeneity, computed with GLCMs in moving windows of 20U20 pixels, best describe the texture and image variability. This is consistent with previous analyses performed on similarly high-resolution sidescan sonar imagery, like DSL-120 data (1-m resolution : Blondel et al, 1993) or TOBI data (6-m resolution : Blondel, 1996). The technique and its results are detailed in 4.…”

Section: Spatial Statistics and Image Analysissupporting

confidence: 90%

“…Textures can be quanti¢ed in di¡erent ways, but theoretical work by Haralick (1979) and practical applications to sonar imagery by Reed and Hussong (1989), Blondel et al (1993), Blondel (1996,2000), Gao et al (1998) and others have shown that co-occurrence matrices are the most adapted tools. Co-occurrence matrices address the average spatial relationships between pixels of a small region.…”

Section: Textural Analysismentioning

confidence: 99%

See 1 more Smart Citation

Textural analyses of sidescan sonar imagery from two mound provinces in the Porcupine Seabight

Huvenne

Blondel

Henriet³

2002

Marine Geology

Self Cite

109

View full text Add to dashboard Cite

Large mound structures have been discovered in the Porcupine Seabight (Northeast Atlantic) at a depth of 5001 200 m, associated with the growth of cold-water deep-sea coral species such as Lophelia pertusa (L.) or Madrepora oculata (L.). During the Training Through Research cruise in 1997, high-resolution OREtech sidescan sonar data were acquired over two provinces of these structures. This article focuses on the presentation and quantitative interpretation of representative sections of these sidescan data from areas around the Belgica and Hovland mounds. Several image analysis tools were used, but texture analysis, based on grey level co-occurrence matrices, gave the best results. Entropy and homogeneity indices were calculated, and the resulting images made it possible to discriminate between different seabed features on a quantitative basis. Mounds, moats and background sediments could be delineated accurately, and the image textures could be linked to the actual seafloor appearance through core descriptions and deep-towed video data. A major difference was found in the acoustic returns from the two provinces studied: the Belgica province shows much rougher textures. This is due to an actual difference in seabed roughness, caused by a difference in bottom currents and sediment dynamics in the two areas. The combined effect of the northward-directed North Atlantic slope current and superimposed internal waves and tides appears to be much stronger in the Belgica province. The reported difference in current strength might well influence the growth of the deep-water coral species in both areas.

show abstract

Section: Spatial Statistics and Image Analysissupporting

confidence: 90%

Section: Textural Analysismentioning

confidence: 99%

Textural analyses of sidescan sonar imagery from two mound provinces in the Porcupine Seabight

Huvenne

Blondel

Henriet³

2002

Marine Geology

Self Cite

109

View full text Add to dashboard Cite

show abstract

“…The distance and angle for the examined pixels whose spatial relationship was displaced was set to 1 pixel and 0°, 45°, 90°, or 135°. Blondel et al [1993] and Blondel [1996] suggested that two textural features of homogeneity and entropy were sufficient for side-scan sonar interpretation among many textural features [e.g., Haralick et al, 1973]. In addition to these two features, we included a textural feature called correlation in the process of classification to obtain better results, as described later.…”

Section: Classification Using Glcmmentioning

confidence: 99%

“…To complement visual inspection of sonar images, we also provide a seafloor classification system generated from textural analysis by means of grey level cooccurrence matrix (GLCM) using the calibrated sonar data. This technique has been successfully applied for marine geophysical and geological surveys and habitat mapping [e.g., Reed and Hussong, 1989;Gao et al, 1998;Blondel et al, 1993;Blondel, 1996Blondel, , 2000Hurst and Karson, 2004;Mitchell and Clarke, 1994;Cochrane and Lafferty, 2002;Huvenne et al, 2002]. In this study, calibrated backscattering levels were effectively incorporated in the classification, which was often not possible in previous studies.…”

Section: Introductionmentioning

confidence: 99%

High‐resolution acoustic mapping to understand the ore deposit in the Bayonnaise knoll caldera, Izu‐Ogasawara arc

et al. 2015

View full text Add to dashboard Cite

We collected deep‐sea multibeam, side scan, and subbottom profiler data using an autonomous underwater vehicle at the Bayonnaise knoll, a submarine caldera located in the rift zone of the Izu‐Ogasawara arc. We aimed to reveal topographic and geological features and the origin of a hydrothermal field called the Hakurei site in the caldera. We performed seafloor classification by textural analysis using calibrated side‐scan sonar data, which provided an effective means to understand the geology and to highlight potential areas of hydrothermal constructions. The high‐resolution bathymetric map illustrates that the Hakurei hydrothermal field is distributed over a landslide landform in the caldera wall. The distribution of hydrothermal vents indicates that the slip surface has served as a major route of hydrothermal fluids. The radial alignment of chimneys and mounds indicates radial routes of hydrothermal fluid and/or belching along fragile lines in the landslide landform. Various postcaldera activities are inferred including the formation of a lava dome, a pyroclastic cone, and subsequent phreatic explosions. A general volcano‐tectonic structure extending across the caldera in a NW‐SE direction is interpreted as an inferred boundary fault of the North Myojin Rift. Analogous to the Hokuroku basin and land kuroko deposits, it is suggested that the main contributing factor in the formation of kuroko deposits was volcano‐tectonic activity that dominated the margin of the back‐arc rift basin. The intersections between the margin of a rift basin and the surrounding knolls have a high potential for ore‐forming areas.

show abstract

“…(1) short-run emphasis (SRE) This concept is also known by other names, such as greytone spatial-dependence matrices (Haralick et al 1973) and grey-level co-occurrence matrices (Weszka et al 1976;Pace & Dyer 1979;Reed & Hussong 1989;Blondel et al 1993). Haralick et al (1973) considered an image of nxn pixels quantised to M grey levels.…”

Section: Grey-level Run-length Featuresmentioning

confidence: 99%

Seabed classification of the South Tasman Rise from SIMRAD EM12 backscatter data using artificial neural networks

Müller

Overkov

Royer³

et al. 1997

Australian Journal of Earth Sciences

View full text Add to dashboard Cite

Textural analysis and structure-tracking for geological mapping: applications to sonar images from Endeavour Segment, Juan de Fuca Ridge

Cited by 13 publications

References 4 publications

Textural analyses of sidescan sonar imagery from two mound provinces in the Porcupine Seabight

Textural analyses of sidescan sonar imagery from two mound provinces in the Porcupine Seabight

High‐resolution acoustic mapping to understand the ore deposit in the Bayonnaise knoll caldera, Izu‐Ogasawara arc

Seabed classification of the South Tasman Rise from SIMRAD EM12 backscatter data using artificial neural networks

Contact Info

Product

Resources

About