RodSizer turns electron-microscopy images into reviewed nanoparticle measurements and size distributions. It brings detection, calibration, particle selection, and reporting into one local application.
Upload TEM, STEM, or SEM images, calibrate the scale, and inspect the detected particles. Selections can be adjusted before generating measurements and combining results across an image folder.
Calibration comes from embedded image metadata, scale-bar OCR, or a manually drawn reference. The review interface lets users inspect particle boundaries and exclude aggregates or ambiguous detections.
The image-analysis pipeline uses K-means thresholding to identify foreground particles and distance-transform watershed to separate touching regions. Rotated bounding boxes provide length, width, aspect ratio, and orientation; additional measurements describe particle area and shape.
Volume is estimated with a hemispherically capped cylinder model. Segmentation builds on AutoDetect-mNP (JACS Au, 2021, 1, 316–327), with watershed splitting and OCR calibration integrated into the measurement workflow.
RodSizer produces particle overlays, per-particle measurements, size distributions, and folder-level statistics. CSV, Excel, and PDF exports carry the reviewed selections and calibration into subsequent analysis.
Shi Chen and Arda Turk — Murphy Group, UIUC.