Adapt Older OpenCV Contour-Based ODLC Solution for New Mission Parameters

[OrangePear13]

Adapt Older OpenCV Contour-Based ODLC Solution for New Mission Parameters

Problem

Adapt code from the old SUAS repository for use in identifying objects that will be in SUAS 2025.

Solution

The old SUAS repository contains code that uses OpenCV contours to locate and classify different shapes. The new SUAS rules have objects instead of shapes, but does not require the specific classification of different shapes.

Parts of the old code can be adapted to identify objects.

The new solution should:

• Have more aggressive image processing
• More strict filtering of contours
• Less reliance on detail being preserved in object contours

Additional Information

The directory with old vision code that used OpenCV contours and edge detection: https://github.com/MissouriMRR/SUAS-2023/tree/develop/vision/standard_object