Home

Robovision is a Python 3.x library that simplifies computer vision tasks typically implemented by FIRST robotics teams.

Robovision:

• Simplifies the basics and hides the complexity of common OpenCV robot vision tasks.
• Does not aim to do everything or provide every function you might need.
• Is meant to be run on a coprocessor, not the RoboRio.

Robovision has been developed on a mix of MacOS and Linux systems (Ubuntu, Raspbian). While it should work on Windows (particularly with the Anaconda Python environment), this hasn't been tested.

Library components

• Core functions
  • Basic image manipulation
  • Edge detection, flattening, etc.
  • Overlays
• Image stream preprocessing
• Target functions
• Multithreaded video streaming

Basic usage:

An example of the basic use of Robovision is:

import argparse
import cv2
import numpy as np
import robovision as rv

# DEFAULT / STARTING VALUES
LOWER = 60         # HSV lower hue color
UPPER = 100        # HSV upper hue color

target = rv.Target()
target.set_color_range(lower=(LOWER, 100, 100), upper=(UPPER, 255, 255))


def main():
    ap = argparse.ArgumentParser()
    ap.add_argument("-s", "--source", required=True,
                    help="Video source, e.g. the webcam number")
    args = vars(ap.parse_args())
    # create a video stream source
    vs = rv.get_video_stream(args["source"])
    # start the source (i.e. turn on the camera)
    vs.start()
    cv2.namedWindow('CapturedImage', cv2.WINDOW_NORMAL)

    while True:
        # read a frame from the camera
        frame = vs.read_frame()
        # get a set of contours around the detected colors in the frame
        contours = target.get_contours(frame)
        # you'd do something here with those contours
        cv2.imshow('CapturedImage', frame)
        # wait for Esc or q key and then exit
        key = cv2.waitKey(1) & 0xFF
        if key == 27 or key == ord("q"):
            cv2.destroyAllWindows()
            vs.stop()
            break