Wrote some utter shit code to get started with tracking via feature detection. Probably not the best way to go about the business.

Click on the screen to pick the green dot to follow

Uses the ORB feature finder. Each feature then has a descriptor and does brute force search to find best one.

Needs a lot more to be useful for anything. Suggestions: reject outliers, take best k matches and pick closest one to previous match, maintain a set of best current descriptors.

This is another approach to following something, but I’m not sure it is a great one.

import cv2
import numpy as np

orb = cv2.ORB()
bf = cv2.BFMatcher(cv2.NORM_HAMMING, crossCheck=True)

minpoint = []
mindes = []
def findNearestKeyPoint(x,y,kp,des):
    global minpoint, mindes
    mindist = 100000000000
    minpoint = kp[0]
    mindes = des[0]
    index = 0
    for point in kp:
        dist = (point.pt[0]-x)**2+(point.pt[1]-y)**2
        if dist < mindist:
            mindist = dist
            minpoint = point
            mindes = des[index]
            #print des[index]
        index += 1

def service_mouse(event,x,y,flags,param):
    global kp, des
    if event == cv2.EVENT_LBUTTONDOWN:

cap = cv2.VideoCapture(0)



    _, frame = cap.read()

    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    kp = orb.detect(gray,None)
    kp, des = orb.compute(gray, kp)
    img2 = cv2.drawKeypoints(frame,kp,color=(0,255,0), flags=0)
    if minpoint:
        cv2.circle(img2, (int(minpoint.pt[0]),int(minpoint.pt[1])) , 4, [0,0,255],-1 )

        matches = bf.match(np.array([mindes]),des)
        # suggestions: get list of best, pick closest to knwon position, update descriptors

        matches = sorted(matches, key = lambda x:x.distance)
        newguy = kp[matches[0].trainIdx].pt
        cv2.circle(img2, (int(newguy[0]),int(newguy[1])) , 10, [0,0,255],-1 )


    k = cv2.waitKey(5) & 0xFF
    if k == 27: