Optimizing your workflow in Agisoft StereoScan (an early, dedicated stereo-photogrammetry tool by Agisoft, whose core algorithms have evolved into Agisoft Metashape) focuses on minimizing computation time while maximizing the reconstruction quality of your 3D model.
Here is how to streamline and optimize each stage of the process: 1. Pre-Processing & Image Selection
Before importing photos into the software, clean your dataset to prevent processing errors.
Filter out blur: Remove blurry, overexposed, or out-of-focus images, as they distort the software’s spatial correlation.
Ensure high overlap: Maintain a 60% side overlap and 80% forward overlap between successive images so the stereoscopic pairing algorithms can easily identify matching features.
Fill the frame: Avoid massive empty backdrops; ensure your target object occupies most of the photo frame to utilize maximum pixel resolution. 2. Streamlining Image Alignment
The initial camera alignment establishes your sparse point cloud. You can speed this up dramatically with a few adjustments:
Cap the points: In the Align Photos dialog window, reduce the default Key Point Limit (e.g., to 40,000) and Tie Point Limit (e.g., to 4,000). Lowering these constraints speeds up processing without sacrificing overall model geometry.
Use Pair Preselection: If your images contain coordinate metadata or were captured in sequence, enable Generic or Reference preselection to avoid testing every possible image combination. 3. Sparse Cloud Optimization (The “Magic Wand”)
Leave a Reply