The AMEE-PPI Method to Extract Typical Outcrop Endmembers from GF-5 Hyperspectral Images

TL;DR

The AMEE-PPI method improves endmember extraction from GF-5 hyperspectral images by combining spatial and spectral information, leading to more accurate geological analysis.

Contribution

AMEE-PPI is a novel hybrid algorithm that integrates AMEE and PPI to enhance endmember extraction accuracy and stability in hyperspectral imagery.

Findings

AMEE-PPI achieved the lowest Spectral Angle Distance (SAD) and Spectral Information Divergence (SID) values across all outcrop types.

The method avoids vegetation endmember leakage and provides cleaner, more representative endmembers compared to existing methods.

AMEE-PPI outperformed PPI, OSP, VCA, and AMEE in accuracy and robustness on GF-5 hyperspectral images.

Abstract

AMEE-PPI combines the AMEE and PPI approaches for improved endmember extraction.

AMEE-PPI outperforms PPI, OSP, VCA, and AMEE in endmember extraction accuracy.

AMEE-PPI yields outcrop endmembers for geological exploration and spectral analysis.

What are the main findings?

A hybrid algorithm named AMEE-PPI was proposed by integrating Automated Morphological Endmember Extraction (AMEE) and Pure Pixel Index (PPI), effectively overcoming limitations of each method and enhancing the precision and stability of endmember extraction from GF-5 hyperspectral images. The algorithm dynamically calculates pixel purity by running PPI within morphological structural elements, thus incorporating both spectral and spatial information.Experimental results on GF-5 hyperspectral images in a geologically complex outcrop region demonstrated that AMEE-PPI achieved superior performance compared to four classical algorithms (PPI, OSP, VCA, and AMEE), with the lowest Spectral Angle Distance (SAD) and Spectral Information Divergence (SID) values across all outcrop types. The extracted endmembers more closely matched ground-truth spectra, significantly improving hyperspectral representation of pure land-cover classes.

A hybrid algorithm named AMEE-PPI was proposed by integrating Automated Morphological Endmember Extraction (AMEE) and Pure Pixel Index (PPI), effectively overcoming limitations of each method and enhancing the precision and stability of endmember extraction from GF-5 hyperspectral images. The algorithm dynamically calculates pixel purity by running PPI within morphological structural elements, thus incorporating both spectral and spatial information.

Experimental results on GF-5 hyperspectral images in a geologically complex outcrop region demonstrated that AMEE-PPI achieved superior performance compared to four classical algorithms (PPI, OSP, VCA, and AMEE), with the lowest Spectral Angle Distance (SAD) and Spectral Information Divergence (SID) values across all outcrop types. The extracted endmembers more closely matched ground-truth spectra, significantly improving hyperspectral representation of pure land-cover classes.

What is the implication of the main finding?

The AMEE-PPI algorithm offers a more robust and accurate approach for endmember extraction in hyperspectral imagery, which is crucial for improving the quality of spectral unmixing, material classification, and object detection in remote sensing applications.By accurately identifying typical outcrop endmembers, the proposed method provides valuable spectral references for geological mapping, mineral exploration, and environmental monitoring, particularly in regions with complex surface compositions.

The AMEE-PPI algorithm offers a more robust and accurate approach for endmember extraction in hyperspectral imagery, which is crucial for improving the quality of spectral unmixing, material classification, and object detection in remote sensing applications.

By accurately identifying typical outcrop endmembers, the proposed method provides valuable spectral references for geological mapping, mineral exploration, and environmental monitoring, particularly in regions with complex surface compositions.

Mixed pixels remain a central obstacle to reliable endmember extraction from hyperspectral imagery. We present AMEE–PPI, a hybrid method that embeds the Pure Pixel Index (PPI) within morphological structuring elements and propagates spectral purity via dilation/erosion, thereby coupling spatial context with spectral cues while avoiding a user-fixed number of projections. On GaoFen-5 (GF-5) AHSI data from a geologically complex outcrop region, we benchmark AMEE–PPI against four widely used algorithms—PPI, OSP, VCA, and AMEE. The pipeline uses HySime for noise estimation and signal-subspace inference to set the endmember count prior to extraction and applies morphological elements spanning 3 × 3 to 15 × 15 to balance spatial support with local heterogeneity. Quantitatively, AMEE–PPI achieves the lowest spectral angle distance (SAD) for all outcrop types—purple–red: 0.135; yellow–brown: 0.316; gray: 0.191—surpassing the competing methods. It also attains the lowest spectral information divergence (SID)—purple–red: 0.028; yellow–brown: 0.184; gray: 0.055—confirming superior similarity to field reference spectra across materials. Visually, AMEE–PPI avoids the vegetation endmember leakage observed with several baselines on purple–red and gray outcrops, yielding cleaner, more representative endmembers. These results indicate that integrating spatial morphology with spectral purity improves robustness to illumination, mixing, and local variability in GF-5 imagery, with direct benefits for downstream unmixing, classification, and geological interpretation.

Figures