A graphical shape based esolang

Web-based Shape & MNIST Recognition Deep Learning Application.

An application to recognise hand drawn shapes using OpenCV 3.4.4

Matlab and Python code to compute perturbed topological signatures (PTS), an efficient topological representation that lies on the Grassmann manifold.

Draw magical symbols to cast spells

The $1 Unistroke Recognizer, a 2D single-stroke recognizer ported to Flutter/Dart

I designed an angular app to identify 8 different shapes using 700 training set of shaps data set which are drawn on html5 canvas and a model trained in keras with tensorflow and python ,to predict in angular app used tensoflowjs.

Currently in research phrase. This project uses OpenCV, tesseract-ocr and other techniques to indentify different types of valves on P&ID (Piping and Instrumentation Diagrams), extract valves's KKS symbols and diameters.

A parallel algorithm for shape recognition implemented in C++ and OpenMP

A MATLAB program that uses k-means clustering to find and classify user types as an extension of a shape recognition research study. Originally developed at Occidental College from October-December 2019.

depth frame analysis with kinect camera and python

This repository is a collection of fundamental digital image processing operations and algorithms performed on greyscale images, or Portable Grey Map (PGM) files, using different data structures in C++, as part of an assignment and final project module for the Data Structures (CS2001) course.

Extracting the coordinates of shapes ndoes from a learning dataset of images. Project for course Machine Learning at University of Muenster.

Shape recognition with the KNN algorithm

Simple shapes recognition app.

A neural network model based on TensorFlow that predicts shape of brick

p5js ML-powered sandbox.

Python-приложение с нейросетью на sklearn, которая в реальном времени распознает нарисованные фигуры (круг, треугольник, квадрат) с оценкой в процентах. Интерфейс на pygame, обучение на 1000+ ручных изображениях, работа с картинками через Pillow, векторные операции через numpy, сохранение модели в .h5.