Stereo Dust Detection and Automatic Threshold Optimizer

John Deere
Spring 2021

The Stereo Dust Detection anlgorithm and Threshold Optimizer tool is a feature that I created during my second EDP rotation at John Deere. This work brought a stereo dust detection algorithm that can be tuned using an automatic threshold optimizer tool into our core software. Note: Images and links to come once product ships.

Overview

  • Description: The Stereo Dust Detection algorithm accurately detects dust in different environments and conditions. The threshold optimizer tool is used to automatically adjust the dust threshold for accurate performance between different applications.
  • Results: Both the stereo dust detection algorithm and threshold optimizer tool are merged into our core software and are being used by multiple applications.
  • Impact: Both of these solutions create a resuable solution to the dust detection and threshold optimization problems, which enables faster and more accruate development.


    • Features

    • A Stereo Dust Detection algorithm that performs well in different environments, applications, and lighting conditions that is currently integrated into 2 applications.
    • An Automatic Threshold Optimization tool that enables computer vision algorithms to be tuned to different applications with 90+% accuracy.


      • Resources

      Images and further information to come once product ships.

      Skills

      Skills

      Embedded Software Development 100%

      Computer Vision 80%

      Machine Learning 70%

      Tools

      C++ 100%

      OpenCV 80%

      Python 60%