Title: GrainSpaceW

Potential Application: Fine-grained recognition, domain adaptation, out-of-distribution recognition
1 / 163900
Open Detailed Viewer
Scoring:
 Report a mistake 
Main Information
Description

The dataset acquisition equipment includes three prototypes: P600, G600, and M600. P600 and G600 consist of industrial cameras, grain holding platforms, and lighting sources for illumination. The M600 is a smartphone camera. Wheat grains can be categorized as NORMAL(NOR) and six types of DU grains: FUSARIUM & SHRIVELLED (F&S) grain, SPROUTED(SD) grain, MOULDY (MY) grain, BROKEN (BN) grain, grain ATTACKED by PESTS (AP) and BLACK POINT(BP) grain. Wheat grains are mainly sampled from The United States (USA), Canada (CAN), Australia (AU), Cambodia (KHM), and China (CHN).

Biological Information
  • Species: Wheat
  • Latin name: Triticum aestivum
  • Imaging tissue: Grain
  • Function Information
  • Computer vision tasks: Image classification
  • Potential application: Fine-grained recognition, domain adaptation, out-of-distribution recognition
  • Labeled instances: 163900
  • Label type: Image-level annotations
  • Related traits: -
  • Imaging
  • Total number of images: 163900
  • Image format: PNG
  • Average resolution: 142×202
  • Dataset size: 7.0G
  • Sampling geographic location: United States; Canada; Australia; Cambodia; China
  • Acquisition equipment: Professional-600 (P600) , General-600 (G600) , Mobile-600 (M600)
  • Sampling platform: Controlled environment stationary platform
  • Sensors: RGB (visible light imaging)
  • Current Slide (image properties)