Skip to content

module-vision.md

Latest commit

 

History

History
60 lines (47 loc) · 2.7 KB

module-vision.md

File metadata and controls

60 lines (47 loc) · 2.7 KB
Raw

jsrc vision

OpenCV-based image analysis tools for plant phenotyping: contour extraction, elliptic Fourier descriptors (EFD), and morphological trait calculation.

extract

Extracts target contours from images. Supports thresholding in grayscale, Lab (a-channel), and HSV (s/v-channel). Adjustable parameters include Gaussian blur, morphological open/close operations, and area/aspect-ratio filters. Outputs sub-images sorted by x or y coordinate, optionally with a binary mask.

jsrc vision extract -i sample.png -o extracted/ \
  --channel a --invert --blur 5 --kernel 3 \
  --open-iters 2 --close-iters 2 \
  --min-area-ratio 0.0005 --max-area-ratio 0.8 \
  --min-aspect-ratio 0.1 --max-aspect-ratio 10 \
  --sort-by x --save-mask
  • -i, --input: input image path.
  • -o, --output: output directory.
  • --channel: threshold channel, one of gray,a,b,s,v (default: gray).
  • --invert: invert threshold result.
  • --blur: Gaussian blur kernel size, odd integer (default: 5).
  • --kernel: morphology kernel size (default: 3).
  • --open-iters: open iterations (default: 2).
  • --close-iters: close iterations (default: 2).
  • --min-area-ratio: minimum contour area ratio (default: 0.0005).
  • --max-area-ratio: maximum contour area ratio (default: 0.8).
  • --min-aspect-ratio: minimum aspect ratio (default: 0.1).
  • --max-aspect-ratio: maximum aspect ratio (default: 10.0).
  • --sort-by: output order by x or y (default: x).
  • --save-mask: save binary mask image.

efd

Computes elliptic Fourier descriptors from extracted contours, compressing shape information into a set of harmonic coefficients. Coefficients can be used for shape clustering or classification. Supports batch processing of .npy contour files, with optional reconstruction preview plots.

jsrc vision efd -i extracted/ -o descriptors/ \
  --harmonics 20 --points 300 --no-plot
  • -i, --input: input .npy file or directory.
  • -o, --output: output directory.
  • --harmonics: number of EFD harmonics (default: 20).
  • --points: reconstruction points for preview (default: 300).
  • --no-plot: skip preview plot generation.

traits

Extracts morphological traits in one step. Given an image, it segments via thresholding and computes contour area, perimeter, dimensions, circularity, eccentricity, and other traits. Convenient for batch morphometric analysis.

jsrc vision traits -i sample.png --channel a --invert --blur 5 --kernel 3
  • -i, --input: input image path.
  • --channel: threshold channel, one of gray,a,b,s,v (default: gray).
  • --invert: invert threshold.
  • --blur: Gaussian blur kernel size, odd integer (default: 5).
  • --kernel: morphology kernel size (default: 3).