Skip to main content

HV Explorer

The HV Explorer is a powerful, Python-based GUI developed by qtec for the visualization, manipulation, and analysis of Hyperspectral Imaging (HSI) datacubes. Designed primarily for the "Proof of Concept" and exploratory phases of a project, it allows researchers and engineers to validate HSI workflows without writing a single line of code.

Visualization of the spectral curves of meat and fat regions in HV Explorer img70p

Visualization of the spectral curves of meat and fat regions in HV Explorer

Under Development

The HV Explorer is under constant development. Since the initial Alpha (March 2025) and the Beta release (June 2025) — which introduced basic ML classification — we have recently focused on quality-of-life improvements and bug fixes (March 2026) in order to ensure a stable foundation. Our next milestone is the much anticipated integration of camera capture functionality.

Current Status:

  • Active Development: We are currently implementing Live Data Capture. This will allow direct integration with the Hypervision cameras, moving the HV Explorer from a post-processing analysis tool to a real-time capture solution.

  • Next on the Roadmap: Once camera integration is stable, we will prioritize the ability to save, load and export ML models.

  • Planned Features: Project management, workflow saving, dead pixel removal and band math remain on our long-term roadmap.

The first stable release (v1.0), featuring live data capture support from the Hypervision cameras, is targeted for June 2026.

System Requirements

  • OS: Windows 11 or Linux (Ubuntu 22.04+ recommended)

  • Environment: Python 3.10 or higher

  • RAM: 8GB Minimum (16GB recommended for large HV1700 cubes)

Key Features​

Data Visualization & Comparison​

  • Format & Interleave Agility: Native support for for importing and exporting PAM, ENVI and TIFF files. The underlying SDK manages BIP, BIL, and BSQ interleave types interchangeably, ensuring seamless data handling across different HSI standards.

  • Spectral Slicing & RGB Composition: Visualize individual spectral bands with customizable color maps or generate false-color RGB images to highlight specific chemical signatures.

  • Multi-Cube Analysis: Open and compare multiple high-resolution images simultaneously — a task made possible by the underlying HV SDK.

Spectral Analysis & Transformation​

  • Flexible ROI Spectroscopy: Plot and compare the mean spectra of multiple Regions of Interest (ROI). The GUI provides flexible selection tools, supporting rectangular, elliptical, and multi-point (free polygon) shapes to precisely isolate features of interest regardless of their geometry.

    • Data Export: Export selected spectral data to CSV or export ROI annotations for use in external analytical pipelines.

  • Preprocessing: Built-in tools for Reflectance Calibration (white/black references), SNV (Standard Normal Variate) normalization, and spectral derivatives (1st1^{st} and 2nd2^{nd} order).

  • Smoothing: Noise reduction using Gaussian or Savitzky-Golay filters.

  • Data Reduction: Perform spatial and spectral cropping and binning in all three dimensions (x,y,λx, y, \lambda) to optimize data cubes for specific analysis tasks.

Advanced Processing & Machine Learning​

The HV Explorer provides an accessible interface for complex HSI transformations and predictive modeling:

  • Dimensionality Reduction: Principal Component Analysis (PCA).

  • Unsupervised Learning: Clustering algorithms (e.g., K-Means) for automated feature discovery.

  • Machine Learning: Build and apply classification and regression models in real-time. Currently supported models include:

    • Classification: Support Vector Machine, Partial least-squares, and Regularized Least Squares.
    • Regression: Support Vector Regression and Least Squares.

Performance: Powered by HV SDK​

The HV Explorer is more than just a GUI; it is a visual wrapper for the HV SDK.

  • The 8x Memory Expansion: Hyperspectral cubes are massive — a raw cube from the HV1700 often exceeds 1GB in uint8uint8 format. When using these cubes in conventional processing tools, memory usage can grow rapidly when performing just a few operations on the initial data. Type conversions to double-precision floats cause each cube to take 8x the amount of memory, transformations such as matrix-multiplication typically double memory usage temporarily, and in general, a lot of processing time is wasted moving and copying memory. In contrast, the HV Explorer utilizes the SDK's lazy pipeline architecture to process only small slices of the cube at a time and only pull the necessary data required for a given operation, leaving RAM usage low and CPU utilization high.

  • Automatic Interleave Optimization: Different HSI operations require different memory access patterns to be efficient. Under the hood, the HV SDK is used to perform seamless and efficient interleave transformations between BIP, BIL, and BSQ depending on the task (spatial viewing vs. spectral plotting), it ensures localized memory access and fluid performance.

  • Multi-Cube Stability: Because of this memory-mapped approach, users can open and compare multiple large datasets simultaneously on a standard workstation without exhausting system resources.

Workflows​

Basic​

Advanced​

Support​

For reporting bugs or requesting assistance write an email to: support@qtec.com

See also the Support section.