Landsat 8 Cloud Detection Using Fractal Thresholding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional cloud detection methods for Landsat 8 snow-containing images face limitations due to similar spectral characteristics of clouds and snow, leading to unreliable detection effects, especially in regions with varying cloud types, snow thickness, and underlying surface types.
Innovation Solution
A cloud detection method utilizing the green waveband, first infrared waveband (1.6 μm), and second infrared waveband (2.1 μm) as principal components, with a fractal summation model to calculate a cloud threshold, followed by hotspot analysis and standard deviation ellipse to remove false anomalies, enhancing precision and universality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional cloud detection methods are used on Landsat 8 snow-containing images, then cloud detection can be performed, but misjudgment occurs due to similar spectral characteristics of clouds and snow
Solution Approach 1:
The patent segments the cloud detection process into multiple independent modules: spectral difference analysis module, texture feature analysis module, and spatial distribution analysis module. Each module processes specific characteristics separately and their results are integrated to achieve accurate cloud-snow differentiation, avoiding the misjudgment caused by similar spectral characteristics.
Solution Approach 2:
The patent transitions from single-dimension spectral analysis to multi-dimensional analysis by incorporating texture features (gray-level co-occurrence matrix), spatial distribution patterns, and spectral differences across multiple wavebands. This dimensional expansion enables reliable cloud detection even when spectral characteristics are similar.
2Adaptability or versatility
If F-mask algorithm with scene-based threshold is applied, then cloud, cloud shadow and snow can be recognized, but the same threshold cannot detect cloud in images with special surface reflectivity
Solution Approach 1:
The patent applies local quality by using adaptive thresholds calculated separately for different regions of the image based on local statistical characteristics. Instead of a single global threshold, each pixel's cloud detection threshold is determined by its local neighborhood properties, enabling accurate detection across varying surface reflectivity conditions.
Solution Approach 2:
The patent implements dynamic threshold adjustment where detection parameters adapt to local image characteristics. The algorithm dynamically calculates thresholds based on local mean and standard deviation, allowing the detection system to automatically adjust to different surface types and reflectivity conditions within the same image.
3Measurement precision
If texture features and gray-level co-occurrence matrix are used for cloud-snow differentiation, then detection accuracy improves, but computational efficiency decreases
Solution Approach 1:
The patent extracts only the most discriminative texture features using gray-level co-occurrence matrix calculations focused on specific directional relationships. By selecting key texture parameters (contrast, correlation, energy, homogeneity) rather than processing all possible features, the method maintains high differentiation accuracy while reducing computational burden.
Data Source
AI summary
A cloud detection method based on Landsat 8 snow-containing image, including the following steps: Step 1, selecting any Landsat 8 image as a current image; Step 2, obtaining a cloud threshold for delineating a cloud range from the current image; and Step 3, removing false anomalies in the cloud range delineated by the cloud threshold from the current image so as to obtain a cloud image from which the false anomalies have been removed. The present disclosure can effectively solve the problem of confusion of cloud and snow present in conventional cloud detection methods, and is applicable to regions of different latitudes, without limitations by the amount of cloud.