Make Computers See: Image Processing with OpenCV

Project Abstract

[Chinese ver. 中文版]

MEMS gyroscopes are a type of sensor that is widely used in phones, cars, and many other devices to detect the rate of rotation. However, the mass production units of the gyroscope may be imprecise due to errors in production. One may need to fine-tune the gyroscopes to achieve higher precision. It can be accomplished by precisely etching parts of it with a laser. To automate the process, photos of the gyroscope will be taken with a microscope camera; a computer program will then recognize parts to be fixed and calculate the actual area to be fixed from the image taken; another program will use the results of the previous program and control the laser etching machine to etch the gyroscope. This project will focus on creating the computer program that recognizes and calculates the parts in the gyroscope to be fixed, and it will integrate with other colleagues’ programs to achieve the overarching goal. OpenCV is an image processing library that will be used in this project to process the image, and the processing algorithm will be implemented with the Python programming language. The major process of the project is to find the contours of parts in the gyroscope; identify certain shapes that make up the gyroscope; calculate the parts to be etched according to the image scale.

Weekly Reports

Week 1 - Intro to the project

Week 2 - Image Preprocessing

Week 3 - Corner Detection

Week 4 - Improving Corner Detection Results

Week 5 - Investigating Camera Calibration

Week 6 - Implementing Camera Calibration