Image classification for grain quality grading: a labelled dataset of pulse varieties and grades (urad, moong and others) with the Python pipeline that processes and classifies them.

---

What this is

A computer-vision project for grading grain quality. The bulk of the repository is the labelled image dataset, organised into one directory per variety-and-grade combination such as urad_med and moong_good, with Python modules handling the image processing and classification over them. 221 tracked files, most of them training images.

2	1
Python modules	docs

---

Architecture

flowchart TB
    R["Repository"]
    N0["urad_med/"]
    R --> N0
    N1["moong_good/"]
    R --> N1
    N2["toor_med/"]
    R --> N2
    N3["rice_bad/"]
    R --> N3
    N4["rice_med/"]
    R --> N4
    N5["toor_bad/"]
    R --> N5

Loading

_{Derived from the directories present in this repository.}

---

Project structure

Path	Contents	Files
urad_med/	—	22
moong_good/	—	19
toor_med/	—	19
rice_bad/	—	18
rice_med/	—	18
toor_bad/	—	18
urad_good/	—	18
rice_good/	—	17
urad_bad/	—	17
channa_bad/	—	16
channa_good/	—	13
toor_good/	—	10
channa_med/	—	7
moong_bad/	—	3

_{221 tracked files in total.}

---

_{gowtham-creator/Multi-Grain-Image-Classification-and-processing}