perseo-quality 1.1.0

Python Ecosystem for Remote Sensing & Earth Observation: SAR Products Quality analyses.

perseo-quality

Python Ecosystem for Remote Sensing & Earth Observation (PERSEO) QUALITY package for SAR products quality analysis and calibration assessment.

Features

Point Target Analysis

Analysis of point targets (passive corner reflectors or transponders) in SAR scenes.

• Impulse Response Function (IRF): range and azimuth resolution, PSLR, ISLR, SSLR (1D and 2D)
• Radar Cross-Section (RCS): RCS estimation, RCS errors, peak phase error, clutter, SCR
• Localization Errors: slant range and ground localization error, azimuth localization error

Radiometric Analysis

Global quality assessment on homogeneous distributed targets.

• Block-wise: automatic scene partitioning into azimuth blocks (bursts for TopSAR/ScanSAR)
  • Noise Equivalent Sigma-Zero (NESZ) profiles
  • Average Elevation Profiles
  • Scalloping Profiles
  • KPI estimation
  • Configurable block size, range margin, outlier removal and smoothing
• Point-wise: profiles extracted around a selected location

Interferometric Analysis

Coherence analysis from interferometric SAR products.

• 2D coherence intensity histograms along range and azimuth
• Burst-by-burst processing with configurable partitioning
• Graphical representation

Spectral Analysis

Spectral content investigation in the frequency domain.

• Point Target: absolute and phase spectra at each target location
• Distributed Target: spectral amplitude on bursts or azimuth blocks
• Range and azimuth profiles at each third of the data portion

Elevation Notch Analysis

Antenna pointing calibration from dedicated Elevation Notch (EN) products.

• Mis-pointing angle estimation by matching measured range profiles with theoretical EAP
• Robust estimation using EN patterns with central low-power "hole"

Target Ambiguity Ratio (PTAR / DTAR)

Signal-to-ambiguity ratio computation.

• PTAR: Point Target Ambiguity Ratio
• DTAR: Distributed Target Ambiguity Ratio
• Left and right ambiguity localization

Data Model

All analyses are designed around a generic Python protocol, making them format-agnostic and independent of the input product type. The protocol and its utilities are available in the perseo_quality.io module.

Installation

pip install perseo-quality[graphs]

The [graphs] extra enables graphical output (matplotlib). For development:

pip install perseo-quality[dev,test,docs]

License

This project is licensed under the MIT License. See the LICENSE.txt file for details.

Copyright © 2026-present Aresys S.r.L. [e-mail]