An OpenCV-based labeling software to generate segmentation and ellipse-fit labels for frames in a video.
More information can be found here.
While installing OpenCV is generally a complex task, Docker can be used to run a minimized Ubuntu system for added portability.
More information can be found here.
The neural network exports numpy binary data in append mode. Reading in this data is not straight forward, so this code generalizes python readability of the binary data.
More information can be found here.