example1.py

#**********************************************************************
#Finds the 100 best features in an image, and tracks these
#features to the next image. Saves the feature
#locations (before and after tracking) to text files and to PPM files, 
#and prints the features to the screen.
#**********************************************************************

#include "pnmio.h"
from __future__ import print_function
from klt import *
from PIL import Image
from selectGoodFeatures import *
from writeFeatures import *
from trackFeatures import *
import pickle, time

def main():
	tc = KLT_TrackingContext()
	nFeatures = 50
	tc.nSkippedPixels = 0
	tc.max_residue = 10.0
	
	KLTPrintTrackingContext(tc)

	img1 = Image.open("img0.pgm")
	img2 = Image.open("img1.pgm")
	ncols, nrows = img1.size

	fl = KLTSelectGoodFeatures(tc, img1, nFeatures)

	print("\nIn first image:")
	for i, feat in enumerate(fl):
		print("Feature #{0}: ({1},{2}) with value of {3}".format(i, feat.x, feat.y, feat.val))

	KLTWriteFeatureListToPPM(fl, img1, "feat1.ppm")
	#KLTWriteFeatureList(fl, "feat1.txt", "%3d")

	#pickle.dump(tc, open("context.dat","w"))
	#pickle.dump(fl, open("featurelist.dat","w"))

	#img1 = Image.open("img0.pgm")
	#img2 = Image.open("img1.pgm")

	#tc = pickle.load(open("context.dat"))
	#tc.writeInternalImages = True
	#KLTPrintTrackingContext(tc)

	#fl = pickle.load(open("featurelist.dat"))

	#KLTTrackFeatures(tc, img1, img2, fl)
	count = 0
	ti = time.clock()
	for i in range(100):
		KLTTrackFeatures(tc, img1, img2, fl)
		KLTTrackFeatures(tc, img2, img1, fl)
		count += 2
	print((time.clock() - ti) / count)

	print("\nIn second image:")
	for i, feat in enumerate(fl):
		print("Feature #{0}: ({1},{2}) with value of {3}".format(i, feat.x, feat.y, feat.val))

	KLTWriteFeatureListToPPM(fl, img2, "feat2.ppm")
	#KLTWriteFeatureList(fl, "feat2.fl", NULL) # binary file 
	#KLTWriteFeatureList(fl, "feat2.txt", "%5.1f") # text file 

if __name__=="__main__":
	main()