Scholarlink

<p>The integration of multispectral data processing supported by artificial intelligence-chosen classification methods and field investigations enables us to construct a coherent geological mapping for the Hafafit–Nugrus–Abu Rusheid area. Landsat-8 (OLI) and ASTER image processing were applied to discriminate the various outcropping rock units. Lithological discrimination was based on false color composite (FCC), principal component analysis (PCA), and minimum noise fraction (MNF) and followed by support vector machine (SVM), artificial neural network (ANN), and maximum-likelihood classifier (ML) classifications. Field verification and field studies at accessible sites define the field relations between different rock units. The resulting map scale 1:100,000 comprises ten classes (ten rock units); Migmatite gneiss (class 1), Hornblende-biotite gneiss (class 2), Psammitic gneiss (class 3), Abu Rusheid Paragneiss (class 4), Metagabbro (class 5), Ophiolitic melange (class 6), Metavolcanics (class 7), Mylonitic granite G1 (class 8), Leucogranite G2 (class 9), and Alkali feldspar granite G3 (class 10). The present study successfully discriminated between hornblende-biotite gneiss and psammitic gneiss using Landsat-8 BR (5/7–5/1–4/3) assigned to RGB. Paragneiss and mylonitic granite (G1) have similar reflectance in the area so SVM and ML classification methods effectively enabled to discriminate between those rock units. Leucogranie (G2) exposure was effectively differentiated using ASTER BR (7/6–2/1–4/6). Landsat-8 bands (7–6-1) assigned to (RGB) delineate the metagabbro exposures from the adjacent rock units. ASTER band ratios (7/6–2/1–4/6) discriminate the metavolcanics outcrop successfully. </p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Implementation of multi-optical data, artificial intelligence, and field investigation for geological mapping of Nugrus–Abu Rusheid region, Southeastern Desert, Egypt

  • 期刊论文

  • 发表日期：

    2025年5月15日

查看全文
  • Salama Mohamed Bahr,
  • Ali El-Sayed Omar,
  • Said Mohamed Said,
  • Ali Farrag Osman,
  • Mohamed Salem Kamar,
  • Ali Mohamed A. Abd-Allah

摘要

The integration of multispectral data processing supported by artificial intelligence-chosen classification methods and field investigations enables us to construct a coherent geological mapping for the Hafafit–Nugrus–Abu Rusheid area. Landsat-8 (OLI) and ASTER image processing were applied to discriminate the various outcropping rock units. Lithological discrimination was based on false color composite (FCC), principal component analysis (PCA), and minimum noise fraction (MNF) and followed by support vector machine (SVM), artificial neural network (ANN), and maximum-likelihood classifier (ML) classifications. Field verification and field studies at accessible sites define the field relations between different rock units. The resulting map scale 1:100,000 comprises ten classes (ten rock units); Migmatite gneiss (class 1), Hornblende-biotite gneiss (class 2), Psammitic gneiss (class 3), Abu Rusheid Paragneiss (class 4), Metagabbro (class 5), Ophiolitic melange (class 6), Metavolcanics (class 7), Mylonitic granite G1 (class 8), Leucogranite G2 (class 9), and Alkali feldspar granite G3 (class 10). The present study successfully discriminated between hornblende-biotite gneiss and psammitic gneiss using Landsat-8 BR (5/7–5/1–4/3) assigned to RGB. Paragneiss and mylonitic granite (G1) have similar reflectance in the area so SVM and ML classification methods effectively enabled to discriminate between those rock units. Leucogranie (G2) exposure was effectively differentiated using ASTER BR (7/6–2/1–4/6). Landsat-8 bands (7–6-1) assigned to (RGB) delineate the metagabbro exposures from the adjacent rock units. ASTER band ratios (7/6–2/1–4/6) discriminate the metavolcanics outcrop successfully.