Week 14 · reflection

Week 14 Capture Evidence Log

Your Week 14 log is an input-quality decision.

Do not report whether the map looked good. Stop one step earlier and defend whether the image set deserved processing.

Product-first capture choice

Record:

  • intended product;
  • smallest feature that matters;
  • camera/view strategy;
  • detail-versus-altitude tradeoff;
  • overlap strategy;
  • one lighting or motion concern.

Rejection ledger

Choose four supplied images and classify them:

ImageConditionKeep / review / rejectCoverage consequence

If rejecting an image creates a geometry gap, say so. Do not quietly upgrade the image quality to save the dataset.

Control network

Sketch the ground-control/reference points and one independent checkpoint.

Identify where the network is strongest and where the project has the weakest reference geometry.

Predicted output defect

Before looking at the supplied processing result, write one defect you expect from the image-set evidence.

Then compare your prediction with the result.

Final dataset disposition

Choose:

  • ready for processing;
  • ready with limitations;
  • partial product only;
  • recapture required.

Cite the quality or coverage evidence controlling the decision.

Exit standard

A strong Week 14 log makes one thing obvious: bad input is an engineering finding, not something to hide so the software can finish.