Image Tools provide analytical operations that operate on the full underlying image, using all available instrument bands. Unlike Insight Tools, these operations are not limited to the currently visualized insight and may modify the current result or generate a new one.
Overview #
Image Tools:
- Use all raw bands from the selected image.
- Are independent of the currently visualized insight.
- May either modify the current visualization or create a new insight.
These tools are designed for deeper analysis where full spectral information is required.
Available Tools #
1. Masks #
Masks allow you to filter the visible data by including or excluding specific areas.
You can:
- Add one or more masks to the current insight.
- Combine multiple masks at the same time.
Available mask types:
- Predefined masks (e.g., vegetation, water, clouds, etc.)
- Private masks created by the user (e.g., from thresholding)
Important behavior:
- Masks can only be added in this panel.
- Existing masks that were previously applied cannot be removed here.
Masks do not create a new insight. Instead, they are applied directly to the current visualization, filtering the data accordingly.
2. Distance to Point (Spectral Similarity) #
This tool computes the spectral similarity between a selected point and the rest of the image.
Workflow:
- Right-click on a point in the map covered by the insight and copy its coordinates.
- Paste the coordinates into the tool input.
The system performs a Spectral Angle Mapper (SAM) analysis using the selected point as reference.
Result:
- A new insight is generated.
- The output is typically displayed in grayscale.
- Each pixel represents the spectral distance inverted and normalized from 0 to 1 to the reference point:
- Lower values → lower similarity
- Higher values → higher similarity
This is useful for identifying areas with similar material or spectral characteristics.
3. Principal Component Analysis (PCA) #
PCA reduces the dimensionality of the image while preserving the most relevant variance.
This analysis is applied to all bands of the image.
You can:
- Select how many principal components to visualize.
- Display up to a maximum of 10 components.
Result:
- A set of new insights are generated.
- Each component represents a transformed version of the original data, highlighting different patterns and variability.
PCA is commonly used to:
- Enhance subtle features
- Reduce redundancy across bands
- Support further analysis or interpretation
Summary #
- Image Tools operate on the full image, not just the displayed insight.
- Some tools modify the current visualization (e.g., masks).
- Others generate new insights (e.g., Distance to Point, PCA).
- These tools are best suited for workflows requiring full spectral information.
Each tool is documented in more detail in its corresponding section.