We use cookies to improve your browsing experience and to analyze our website traffic. By clicking “Accept All” you agree to our use of cookies. Privacy policy.
15 reads

VisionMath - AI mathematical expression solver

Table of contents

VisionMath

AI-Powered simple mathematical expression solver

Details

Details on the general process can be found in the PDF presentation and on the JupyterNotebook PDF.

Description

VisionMath takes a handwritten mathematical operation and uses AI to recognize the characters.
Afterward, it builds the operation and returns the corresponding value.

Dependencies

VisionMath was made on Python 3.11 and the packages it requires are:

  • OpenCV
  • Numpy
  • Matplotlib
  • Keras
  • Scipy

Keras Model

VisionMath uses a Tensorflow Keras model formed by:

  1. Flatten Layer (input_shape = (28,28))
  2. Dense Layer (units = 128, activation = relu)
  3. Dense Layer (units = 128, activation = relu)
  4. Dense Layer (units = 13, activation = softmax)

model_image.png

GitHub

Repository can be found here

Tests

Input image

test2.jpeg

Left-Right ordered contours image

test15_contours.png

Model classification

test15_classification.png

Expression evaluation

EXPRESSION: ((4-2)/(6))

RESULT: 0.33