Video processing and Image processing

This part subsamples and divides video using OpenCV. Iterating over video frames and extracting them at a lower frame rate yields a directory of images. This method helps machine learning models process, analyse, and train pictures.

read more →

Sparse Reconstruction

Sparse reconstruction is a major approach used in this project for recovering signals or images from a small or imperfect data collection. The project's goal is to reduce the amount of data points needed to recreate the original signal or image by taking advantage of the data's inherent sparsity.

read more →

Point Cloud Operation

noise reduction, outlier removal, normal estimation, and point cloud alignment.

read more →

Possion and Laplacian

Poisson reconstruction creates a smooth, continuous 3D surface from point cloud data. Poisson reconstruction captures sharp details and complex structures. Modelling the surface as a 3D scalar function's zero-level set uses an approximation of the object's indicator function. The zero-level set from the reconstructed surface is the smooth scalar function after numerically solving the Poisson equation. Poisson reconstruction increases 3D surface quality and fidelity, preparing 3D data interpretation and processing.

read more →