Lucas Kanade Tracker (OpenCV)

LKTrack.py

import cv2
import numpy as np
import itertools

lk_params = dict( winSize  = (10, 10), 
                  maxLevel = 5, 
                  criteria = (cv2.TERM_CRITERIA_EPS | cv2.TERM_CRITERIA_COUNT, 10, 0.03))   

feature_params = dict( maxCorners = 3000, 
                       qualityLevel = 0.5,
                       minDistance = 3,
                       blockSize = 3)

def lk():
    cam = cv2.VideoCapture(0)
    _, img = cam.read()
    oldg = cv2.cvtColor(img, cv2.cv.CV_BGR2GRAY)
    bb = [200,200,300,300]
    old_pts=[]
    while True:
        try:
            _, img = cam.read()
            img1 = img[bb[0]:bb[2],bb[1]:bb[3]]
            #print img1.shape
            g = cv2.cvtColor(img1, cv2.cv.CV_BGR2GRAY)
            #print g.shape
            pt = cv2.goodFeaturesToTrack(g, **feature_params)
            for i in xrange(len(pt)):
                pt[i][0][0] = pt[i][0][0]+bb[0]
                pt[i][0][1] = pt[i][0][1]+bb[1]
            newg = cv2.cvtColor(img, cv2.cv.CV_BGR2GRAY)
            p0 = np.float32(pt).reshape(-1, 1, 2)
            p1, st, err = cv2.calcOpticalFlowPyrLK(oldg, newg, p0, None, **lk_params)
            p0r, st, err = cv2.calcOpticalFlowPyrLK(newg, oldg, p1, None, **lk_params)
            d = abs(p0-p0r).reshape(-1, 2).max(-1)
            good = d < 1
            new_pts = []
            for pts, val in itertools.izip(p1, good):
                if val:
                    new_pts.append([pts[0][0], pts[0][1]])
                    cv2.circle(img, (pts[0][0], pts[0][1]), 5, thickness=3, color=(255,255,0))
            bb = predictBB(bb, old_pts, new_pts)
            if bb[0]+bb[2] >= img.shape[0]:
                bb[0] = img.shape[0] - bb[2] - 1
            if bb[1]+bb[3] >= img.shape[1]:
                bb[1] = img.shape[1] - bb[3] - 1
            old_pts = new_pts
            oldg = newg
            cv2.rectangle(img, (bb[0],bb[1]),(bb[0]+bb[2],bb[1]+bb[3]), color=(255,0,0))
            cv2.imshow("LK", img)
            cv2.waitKey(1)
        except KeyboardInterrupt:
            break
            
def predictBB(bb0, pt0, pt1):
    if not pt0:
        pt0 = pt1
    dx=[]
    dy=[]
    for p1, p2 in itertools.izip(pt0, pt1):
        dx.append(p2[0]-p1[0])
        dy.append(p2[1]-p1[1])
    if not dx or not dy:
        return bb0
    cen_dx = round(sum(dx)/len(dx))
    cen_dy = round(sum(dy)/len(dy))
    print cen_dx, cen_dy,
    print "shift"
    bb = [int(bb0[0]+cen_dx), int(bb0[1]+cen_dy), int(bb0[2]), int(bb0[3])]
    if bb[0] <= 0:
        bb[0] = 1
    if bb[1] <= 0:
        bb[1] = 1
    return bb
lk()

Hey Jay,

I seem to get this issue,

0.0 0.0 shift
-20.0 14.0 shift
-18.0 30.0 shift
-35.0 17.0 shift
-44.0 18.0 shift
-58.0 97.0 shift

Traceback (most recent call last):
File "/media/mathew/d19f4260-bc8b-44b2-9209-1ed9bde9abfd/home/mat/Desktop/dddd.py", line 75, in
lk()
File "/media/mathew/d19f4260-bc8b-44b2-9209-1ed9bde9abfd/home/mat/Desktop/dddd.py", line 28, in lk
for i in xrange(len(pt)):
TypeError: object of type 'NoneType' has no len()

I have the same error