Demo project with seeds photos for research. Can be used as a playground for:
- testing different models: detection / segmentation / instance segmentation
- demo for synthetic data
- all images have the same resolution: 4032 * 2268 => so it is a good test for sliding window approaches
- modeling some close tasks from different industries, images may have similar properties in terms of lighting conditions, number of objects and their sizes, camera positions, etc; for example:
- agriculture (find plants diseases, count items, inspect their properties (e.g. size) and quality)
- medicine (x-ray)
- microbiology (microscope images)
- visual inspection and quality control in production (defects, holes, production lanes)
- and so on ...
-
All images have the same resolution 4032 x 2268 x 3 (height x width x channels)
-
There are no labeled objects
-
Datasets:
01_background
- background images without objects02_sunflower
- sunflower seeds with different density03_pumpkin_peeled
- peeled pumpkin seeds with different density04_pumpkin_unpeeled
- unpeeled pumpkin seeds with different density05_pumpkin_mx
- mix of peeled and unpeeled pumpkin seeds with different density06_mix
- mix of sunflower, peeled and unpeeled pumpkin seeds
- Classes