lucas_kanade.py

##This will find velocity vectors of optica flow



import imutils
import numpy as np
import cv2
import time
from time import sleep
from scipy import signal
from PIL import Image
from numpy import *
import random
from matplotlib import pyplot as plt
from pylab import *
def LUCAS_KANADE(image1,image2):
        print('haii')
        I1 = np.array(image1)
        I2 = np.array(image2)
        Shape = np.shape(I1)

        Image1_smooth = cv2.GaussianBlur(I1, (3,3), 0)
        Image2_smooth = cv2.GaussianBlur(I2, (3,3), 0)
        # First Derivative in X direction
        Ix = signal.convolve2d(Image1_smooth,[[-0.25,0.25],[-0.25,0.25]],'same') + signal.convolve2d(Image2_smooth,[[-0.25,0.25],[-0.25,0.25]],'same')
        # First Derivative in Y direction
        Iy = signal.convolve2d(Image1_smooth,[[-0.25,-0.25],[0.25,0.25]],'same') + signal.convolve2d(Image2_smooth,[[-0.25,-0.25],[0.25,0.25]],'same')
        # First Derivative in XY direction
        It = signal.convolve2d(Image1_smooth,[[0.25,0.25],[0.25,0.25]],'same') + signal.convolve2d(Image2_smooth,[[-0.25,-0.25],[-0.25,-0.25]],'same')
        features = cv2.goodFeaturesToTrack(Image1_smooth,8000,0.01,10)
        feature = np.int0(features)
        for i in feature:
                x,y = i.ravel()
                cv2.circle(Image1_smooth,(x,y),3,0,-1)     #image,center,radius,colour of circle,thickness of line           

        u = v = np.nan*np.ones(Shape)
        for l in feature:
                j,i = l.ravel()
                IX = ([Ix[i-1,j-1],Ix[i,j-1],Ix[i+1,j+1],Ix[i-1,j],Ix[i,j],Ix[i+1,j],Ix[i-1,j+1],Ix[i,j+1],Ix[i+1,j-1]])# x-cmpnt of gradient vector
                IY = ([Iy[i-1,j-1],Iy[i,j-1],Iy[i+1,j+1],Iy[i-1,j],Iy[i,j],Iy[i+1,j],Iy[i-1,j+1],Iy[i,j+1],Iy[i+1,j-1]])# y-cmpnt of gradient vector
                IT = ([It[i-1,j-1],It[i,j-1],It[i+1,j+1],It[i-1,j],It[i,j],It[i+1,j],It[i-1,j+1],It[i,j+1],It[i+1,j-1]])#x-y cmpnt of gradient vector


                LK = (IX, IY)
                
                LK = np.matrix(LK)
                
                LK_T = np.array(np.matrix(LK)) 
                
                LK = np.array(np.matrix.transpose(LK)) # transpose of A
                
                A1 = np.dot(LK_T,LK) #Psedudo Inverse
                A2 = np.linalg.pinv(A1)
                A3 = np.dot(A2,LK_T)
                
                #sleep(2)
                (u[i,j],v[i,j]) = np.dot(A3,IT)
                print(v[i,j],u[i,j])
        colors = "rgb"
        
        c=colors[2]
        plt.imshow(I1,cmap = cm.gray)
        for i in range(Shape[0]):
	        for j in range(Shape[1]):
		        if abs(u[i,j])>t or abs(v[i,j])>t: 
			        plt.arrow(j,i,v[i,j],u[i,j],head_width = 5, head_length = 5, color = c)   ##plotting arrays
                                
			
        plt.show()






cap = cv2.VideoCapture('')     ## you can insert any video file
t = 0.3                       ## Threshold for u&v values
while cap.isOpened():
        ret, frame = cap.read()
        if ret:
                ret, frame = cap.read()
                gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
                
                #cv2.imshow('sd',frame)
                #cv2.waitKey(0)
                ret,frame = cap.read()
                gray1 = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
                LUCAS_KANADE(gray, gray1)