Skip to content

summary.md

Latest commit

 

History

History
122 lines (87 loc) · 3.12 KB

summary.md

File metadata and controls

122 lines (87 loc) · 3.12 KB

StarSpikes V2 – Current State Summary

Overview

StarSpikes V2 is a working diffraction spike rendering pipeline for astrophotography images. The system detects stars, filters out nebula/false positives, and applies physically-inspired diffraction spikes with user-adjustable parameters via a GUI.

✅ Completed / Working

1. Environment & Execution

  • Python environment (pyenv + venv) stable
  • All required libraries installed and working:
    • numpy, opencv, astropy, photutils, matplotlib, pillow
  • GUI launches and processes images successfully

2. Star Detection Pipeline

  • Background modeling via Background2D
  • High-pass filtering applied to suppress nebula (Gaussian subtraction)
  • Detection via DAOStarFinder
  • Converted to luminance for more consistent detection

Filtering (multi-stage)

  • Sharpness filter
  • Flux percentile filter
  • Peak percentile filter
  • Roundness constraint

Result

  • Reliable star-only detection
  • Nebula largely excluded from detections
  • Detection sensitivity controllable via UI

3. Diffraction Spike Rendering

  • 4-spike pattern (corrected from 8)
  • Rotation angle implemented and working
  • Spike properties:
    • Length scaled by star brightness (flux)
    • Thickness scaled by brightness
  • Rendering uses overlay + masked blending (prevents nebula blowout)

4. Image Processing / Display

  • FITS → normalized → display-safe image pipeline
  • Proper handling of:
    • dtype conversion (uint16 → uint8)
    • channel layout (C,H,W → H,W,C)
    • OpenCV compatibility (contiguous arrays)

5. Preview System

  • Large preview implemented using matplotlib
  • Uses percentile stretch (1–99%) to avoid oversaturation
  • Supports zoom and pan via matplotlib toolbar
  • Current improvement: should be triggered via GUI button instead of auto-popup

⚠️ Known Limitations / Next Work

1. Spike Appearance (Primary Remaining Work)

Current spikes are:

  • Geometrically correct
  • Visually "drawn" (uniform intensity)

Needed:

  • Radial intensity falloff along spike
  • Optional tapering / soft edges
  • More physically realistic diffraction behavior

2. Detection Tuning

  • High-pass filter kernel size strongly affects detection
  • Percentile thresholds can be too aggressive
  • Need better balance or adaptive thresholds

3. GUI / UX Improvements

Planned:

  • Add "Large Preview" button
  • Optional auto-preview toggle
  • Potential embedded zoom/pan preview (instead of matplotlib window)

4. Output

  • FITS saving not yet implemented
  • No PNG/JPG export yet

🧠 Key Milestones Achieved

  • Transitioned from raw detection → nebula-suppressed star detection
  • Transitioned from debug rendering → physically plausible spike overlay
  • Established stable processing pipeline and UI controls

🚀 Next Steps (Priority Order)

  1. Implement spike intensity falloff (highest impact visual improvement)
  2. Add GUI-controlled large preview button
  3. Fine-tune detection thresholds (possibly adaptive)
  4. Add output saving (FITS + PNG)

Status

✔ Functional V2 pipeline
✔ Visually usable results
➡ Entering refinement / realism phase