Computer Vision Simulation | Object Recognition
Interactive computer vision simulation – learn how object and image recognition works.
A hands-on object detector built from classical image processing: noise filtering, automatic Otsu thresholding, connected-component labelling and shape features, scored against exact ground truth with IoU, precision, recall and F1. All scenes are synthetic, so you can raise the noise or lower the contrast and watch precisely where and why the detector starts to fail.
This simulation demonstrates basic computer vision concepts, such as edge detection and feature extraction. Users can manipulate input images and observe how algorithms identify patterns and shapes within the visual data.
2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install