zeljobers/unikg_cv

(Serbian) Computer vision homeworks and seminary project

Računarski vid

1. CNN klasifikacija nad MNIST dataset-om;

• 2. vežba detekcija ivica.

2. Detekcija kontura gde:
   • Uz 1.jpg generiše se u contours_simple_image1.jpg. CHAIN_APPROX_SIMPLE, RETR_TREE
   • Uz 2.jpg generiše se u contours_image1_complex.jpg. CHAIN_APPROX_NONE, RETR_CCOMP.
3. Dostavljanje frames-ova na kojima su prepoznata lica;
4. Prepoznavanja vrste saobraćajnog znaka;
   • https://www.kaggle.com/datasets/valentynsichkar/traffic-signs-dataset-in-yolo-format
     • yolov3_ts_train_5000.weights mora da se skine odavde
   • https://www.youtube.com/watch?v=GgGro5IV-cs
   • https://www.youtube.com/watch?v=_FNfRtXEbr4

Seminarski rad: Implementacija prebrojavanja entiteta na slici

• https://www.geeksforgeeks.org/count-number-of-object-using-python-opencv/
• https://stackoverflow.com/questions/60957257/count-cells-on-image-using-python-and-opencv
• http://www.sixthresearcher.com/counting-blue-and-white-bacteria-colonies-with-python-and-opencv/

Mentori:

• prof. dr. Igor Saveljić
• prof. dr. Bogdan Milićević