Circuit images generating for 3D printers
Matlab script for the conversion of a circuit pattern photo to a 3d representation which can then be printed on a 3d printer.
The goal of the project was to develop a Matlab script capable of converting a simple photo of a circuit board pattern into a 3D representation that could be printed on a 3D printer.
Conventional image processing methods were used to build the algorithm:
- Background evaluation and detection
- Noise removal
- Simple shape detection (line, circle, rectangle)
- Shape vectorization and smoothing
- .dxf file generation
The main challenge during the development was finding a balance of precision during conversion and avoiding incorporation of noises from the original image into the resulting file.
For example, due to shadows, light conditions, noise on the image the circular slots had small indentations on the borders. It was required to exclude them from printed borders.
At the same time, we were avoiding distorting improvements of an initial image. The solution lied in analyzing properties of shapes in different conditions and generating appropriate criteria for conversion.
- Object detection
- Noise removal
- 3D modeling
Virtual try-on tool for makeup products
The system consists of a face detection and segmentation model and an algorithm that allows recoloring objects without losing their original texture.
Online sign language interpreter
AI algorithm that converts video of a person using sign language into a text transcript
Workout helper app
Mobile app for the estimation of proper body positions during the workout.