Image Processing Lab

Sampling, filtering, Fourier transforms, segmentation, SIFT and HOG, applied to a live image you can change.

About the Image Processing Lab

Every photo filter, scanner and face detector is built from a small set of image-processing operations. The Image Processing Lab applies each one to a live image, so you can see what a blur kernel, a gamma curve or a Fourier filter actually does to the pixels.

The lab starts where a digital image starts, with sampling and quantisation, and moves through contrast and histograms, colour spaces and convolution filters to the frequency domain, noise removal and segmentation. It ends with SIFT keypoints and HOG descriptors, the hand-built features that computer vision relied on before deep learning.

It suits students taking a digital image processing or computer vision course, and anyone curious about what happens inside an editing app. All the processing runs in your browser.

Experiments in this lab

Digitization: Sampling and quantization: how a scene becomes pixels.
Point processing: Negatives, thresholds, gamma and contrast, with the histogram.
Colour spaces: RGB, HSV and greyscale channels side by side.
Spatial filtering: Blur, sharpen and edge kernels slid over the image.
Frequency domain: The Fourier spectrum, and filtering it.
Restoration: Add noise, then remove it with the right filter.
Segmentation: Split the image into regions by threshold and clustering.
SIFT and HOG: Keypoints and gradient histograms used to describe images.

Transformer Lab — Tokenize with BPE, watch attention heads, rotate positions with RoPE, size a KV cache, route tokens to experts and quantize weights.
RAG Lab — Chunk a handbook, search it by meaning, fuse BM25 with dense retrieval, rerank, measure recall and NDCG, and keep the index fresh.
Linear Algebra Lab — Drag vectors to see a span collapse, warp the grid with a matrix, hunt eigenvectors by hand, run Gram–Schmidt step by step, and watch a circle become an ellipse under the SVD.
Machine Learning Lab — Find principal components, run k-means and EM step by step, tame overfitting with ridge and lasso, and compare kNN, trees, naive Bayes, SVMs and boosting.
Deep Learning Lab — Build neural networks from scratch. Explore 3D error surfaces, optimizers, backpropagation, overfitting and convolutions.
All simulators