Skip to content

Commit

Permalink
RGB readme added
Browse files Browse the repository at this point in the history
  • Loading branch information
Sungchan Oh committed Jun 27, 2024
1 parent e12e3e3 commit 4c87d52
Show file tree
Hide file tree
Showing 2 changed files with 132 additions and 0 deletions.
52 changes: 52 additions & 0 deletions README.md
Original file line number Diff line number Diff line change
@@ -1,2 +1,54 @@
# DataExploration-AAPF
Exploratory data analysis for AAPF dataset (RGB, HSI, XRAY)



# Change xlsx files to csv files

# Load csv files
# Merge DFPs? range->text
# Replace DFPs with growth stage if available


# RGB, HSI
# Check input variables with repeatability and ANOVA



# RGB-visualization



# Column header
# frame_nr: frame number
# Angle: side view angle when vegetation is mostly dispersed along the plane
# Width: width measured by green box in MES file (see 2.1)
# Height: height measured by green box in MES file (see 2.1)
# Surface: vegetation area measured by red polygon (see 2.3)
# Convex hull: convex hull area measured by blue polygon (see 2.2)
# Roundedness: roundedness measured by red polygon (unsure, see 2.3)
# Center_of_mass_distance: unclear
# Center_of_mass_x: x coordinate of center of mass point (see 2.4)
# Center_of_mass_y: y coordinate of center of mass point (see 2.4)
# Hue: average hue value
# Saturation: average saturation value
# Intensity: average intensity value
# Fluorescene: average flurescence value
# H##: frequency of pixels with a hue value of ## (##: 0-359)
# S##: frequency of pixels with a saturation value of ## (##: 0-99)
# V##: frequency of pixels with a intensity value of ## (##: 0-99)
# F##: frequency of pixels with a flurescence value of ## (##: 0-99)
#
# Row index (view)
# TOP FRAME: Ignore this row
# TOP AVG: Top view
# SideBottom FRAME 0 to 11: Side view data from various angles (0, 30, ..., 330 degree)
# SideBottom AVG: average of SideBottom FRAME data








80 changes: 80 additions & 0 deletions README_RGB.rm
Original file line number Diff line number Diff line change
@@ -0,0 +1,80 @@

1. Introduction

This document describes the file system structure and conventions used for data collected by the AAPF RGB imaging sensor. The data includes RGB and fluorescence images of plants along with derived morphological measurements and phenotypic information.

1.1. File System Structure

The data is organized into files containing images and spreadsheets based on the experiment, scanning time, and specific plant. Here's an example breakdown:

Parent directory: Named as Experiment_{Experiment_Nubmer}, e.g., Experiment_394
Subdirectories: Named as {Experiment_Number}_{Treatment}, e.g., 394_SandSoybeanHigh


1.2. Filename Conventions
Individual files (interim data products) are located under subdirectories. Filenames follow a specific format that encodes information about the data:

VIS_<Data Type>_<POT_BARCODE>_<Time_In>_<RGB_or_Flurescence>-<View>-<Angle>-<Image_Product_Type>_<Experiment_Number>_<Treatment>_<Time_Out>.png

- VIS: Prefix indicating data from the visible light imaging chamber
- <Data Type>:
- R: Original image
- V: Image with vegetation segments only
- P: Phenotype data (spreadsheet)
- <POT_BARCODE>: Unique identifier for the plant pot
- <Time_In>: Time (YYMMDDHHMMSS.SSS) when the pot entered the chamber
- <RGB_or_Fluorescence>: Type of input sensor image (RGB or FLUO)
- <View>: View from which the image was captured:
- Top: Top-down view
- SideBottom, SideSmall, SideTop, SideFull for the low to tall plants. This corresponds to the camera height at which the sensor is located
- <Angle>: Angle of the camera (0, 30, ..., 330 for side views, 0 for top-down view)
- <Image_Product_Type>: PNG for RGB image, SEG for segmentation results, MES for vegetation masking results augmented with morphological measurement info such as bounding box, convex hull
- <Experiment_Number>: Experiment number within AAPF
- <Treatment>: Treatment info
- <Time_Out>: Time (YYMMDDHHMMSS.SSS) when the pot left the chamber






2. How to Read Image Files Containing Morphological Measurement Info (PNG files containing MES in the filename)

These image files visually represent morphological measurements on the vegetation segments. Color overlays indicate different aspects:
Gray (pixels): vegetation pixels segmented
Green (Polyline): Smallest enclosing rectangle of the vegetation segments
Blue (Polyline): Convex hull (smallest enclosing polygon) of the vegetation segments
Red (Pollyline): Pixel-wise boundary of vegetation segments
Cyan (point): Center of mass coordinate



3. How to Read Information in spreadsheet files (.xlsx for master file, csv file for table)

The spreadsheets contains various phenotypic measurements derived from the RGB images. Here's a breakdown of the column headers:

frame_nr: Frame number from 0-11 for side view data
Angle: Side view angle where the vegetation is most dispersed
Width, Height: Dimensions of the green box in the corresponding MES file (Section 2)
Surface: Area of the vegetation measured by the red polygon (Section 2)
Convex hull: Area of the convex hull measured by the blue polygon (Section 2)
Roundedness: Roundness of the red polygon (calculation method unclear)
Center_of_mass_distance: Unclear meaning
Center_of_mass_x, Center_of_mass_y: Coordinates of the center of mass point (Section 2)
Hue, Saturation, Intensity: Average values for these color properties in the image
Fluorescence: Average fluorescence intensity
H###, S##, V##: Frequency of pixels with specific hue (###: 0-359), saturation (##: 0-99), and intensity (##: 0-99) values
F##: Frequency of pixels with a specific fluorescence value (##: 0-99)

View:
TOP FRAME: Ignore this row
TOP AVG: Data from the top view
Side{Bottom, Small, Top, Full} FRAME 0 to 11: Data from various side view angles (0, 30, ..., 330 degrees)
Side{Bottom, Small, Top, Full} AVG: Average of Side view FRAME data


4. Conclusion

This document provides a guide to navigate raw RGB and phenotypical data tables represented either as masterfile (.xlsx) or individual measurement file (.csv).


0 comments on commit 4c87d52

Please sign in to comment.