Converting rough hand-drawn images to awesome AutoCad-like floor plans.

Creating floor plans, though seemingly simplistic, is a very time-consuming and skill-intensive process. For existing properties, the site officer first makes a visit and then gives a rough drawing (with not-to-scale dimensions marked) and/or some explanation to the designer who must convert this information into a floor plan using his CAD skills. Through this project, we aim to reduce the time spent in the process and the requirement of skilled (Auto)CAD designer by directly converting the rough drawing to a floor plan.

• Python
• Image processing (probably using OpenCV)
• Machine Learning and/or deep learning (potential)
• Programmatically operating AutoCAD (or other similar softwares)

input image

output image

Step 1: Detection of lines and corners
Resources:

• (http://answers.opencv.org/question/25526/corners-and-lines-detection-in-hand-drawn-pictures-in-python/)
• (https://stacks.stanford.edu/file/druid:cg133bt2261/Ahmed_Wheeler_Generating_PowerPoint_Slides_from_Hand_Drawn_Sketches.pdf)

• Knowledge of Python
• Some experience with OpenCV and/or Image Processing is a big plus.

• Utkarsh Mittal - [email protected]

See also the list of contributors who participated in this project.