The Analytical Tools window allows you to apply advanced visual, algorithmic, and spectral adjustments directly to your selected insight. This window specifically modifies the visualization of the selected index you’re currently viewing.

Accessing Analytical Tools #

• Click directly on the index chip you wish to modify.
• The Analytical Tools window will open, providing access to three main categories: Visual, Algorithmic, and Spectral tools.

Visual Tools #

1. Resampling

This setting adjusts how pixel data is aggregated or interpolated:

• Neighborhood (Native): Retains original pixel structure.
• Bilinear: Applies bilinear convolution to smooth pixel transitions.
• Bicubic: Uses bicubic convolution for smoother visual representation.

Note: Resampling settings affect visualization only and do not modify download results.

2. Stretching (Color Scaling)

Adjusts how colors are mapped to your data range:

• Relative methods:
  • 1 Sigma, 2 Sigma, 3 Sigma
  • 90%, 95%, 98%, 100%
• Absolute method: Sets absolute minimum and maximum bounds.

Note: Stretching settings affect visualization only and do not alter downloads.

3. Masks (Only available for instruments supporting masks)

Apply or remove masks such as:

• Vegetation
• Clouds
• Water
• Snow
• Urban
• Custom-defined masks (Depends on AOI)

Important:

• Default masks set in prior settings (not within Analytical Tools) cannot be removed here.
• Masks applied via Analytical Tools will affect both visualization and downloaded results.

4. Palette and Threshold (Only available for indices with predefined color palettes)

• Palette:
  • Default color palette (standard)
  • Grayscale
• Threshold:
  • Set minimum and maximum values to define visibility.
  • Pixels outside this range become transparent.

Note: Palette and Threshold settings affect both visualization and downloaded results.

Algorithmic Tools #

These tools apply advanced processing algorithms:

1. Edge Detection

Apply edge detection to highlight boundaries within your data using algorithms:

• Zero-crossing: Detects edges based on zero-crossings of the second derivative.
• Canny: Provides accurate and clean edge detection using gradient intensity and non-maximum suppression.
• Laplacian: Identifies edges by applying a second-order derivative operator.
• Sobel: Highlights edges by calculating the gradient of pixel intensity.

2. Unsupervised Classification (Clustering)

Group pixels based on spectral similarity:

• K-means: Set the exact number of clusters (classes), grouping data accordingly.
• X-means: Specify the maximum number of clusters; algorithm automatically selects the optimal number of classes based on data structure.

Note: When one algorithmic tool is applied, other algorithmic or spectral tools become unavailable for that insight.

Spectral Tools #

Spectral adjustments and analyses, primarily for hyperspectral data:

1. Distance to Point

• Enter coordinates to perform a Spectral Angle Mapper (SAM) analysis.
• Generates a new insight indicating spectral similarity of each pixel to the selected point.

2. Analyze Range (Only for hyperspectral satellites)

• Select specific wavelength ranges (in micrometers) to analyze:
  • Reflectance
  • Absorption
  • Material abundance (based on absorption features)

Note: Applying spectral tools disables the use of algorithmic tools and vice versa.

Best Practices #

• Use appropriate resampling methods to optimize visual analysis without altering original data.
• Apply stretching techniques to enhance data contrast and interpretation.
• Manage masks carefully, particularly when downloading data, to ensure relevant areas are clearly defined.
• Leverage palette and threshold settings to highlight key areas of interest and improve clarity in downloaded insights.
• Clearly choose the appropriate algorithmic or spectral tools depending on the analytical goals