DT

gistfile1.txt

	# import the necessary packages
from __future__ import print_function
from sklearn.metrics import classification_report
from sklearn.ensemble import RandomForestClassifier
from sklearn.tree import DecisionTreeClassifier
from imutils import paths
import numpy as np
import argparse
import mahotas
import cv2
import sklearn
 
 # handle older versions of sklearn
if int((sklearn.__version__).split(".")[1]) < 18:
	from sklearn.cross_validation import train_test_split
 
# otherwise we're using at lease version 0.18
else:
	from sklearn.model_selection import train_test_split
 

def describe(image):
	# extract the mean and standard deviation from each channel of the image
	# in the HSV color space
	(means, stds) = cv2.meanStdDev(cv2.cvtColor(image, cv2.COLOR_BGR2HSV))
	colorStats = np.concatenate([means, stds]).flatten()
 
	# extract Haralick texture features
	gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
	haralick = mahotas.features.haralick(gray).mean(axis=0)
 
	# return a concatenated feature vector of color statistics and Haralick
	# texture features
	return np.hstack([colorStats, haralick])
	
# construct the argument parser and parse the arguments
ap = argparse.ArgumentParser()
ap.add_argument("-d", "--dataset", required=True, help="path to 8 scene category dataset")
ap.add_argument("-f", "--forest", type=int, default=-1,
	help="whether or not a Random Forest should be used")
args = vars(ap.parse_args())
 
# grab the set of image paths and initialize the list of labels and matrix of
# features
print("[INFO] extracting features...")
imagePaths = sorted(paths.list_images(args["dataset"]))
labels = []
data = []
 
# loop over the images in the input directory
for imagePath in imagePaths:
	# extract the label and load the image from disk
	label = imagePath[imagePath.rfind("/") + 1:].split("_")[0]
	image = cv2.imread(imagePath)
 
	# extract features from the image, then update the list of lables and
	# features
	features = describe(image)
	labels.append(label)
	data.append(features)

# construct the training and testing split by taking 75% of the data for training
# and 25% for testing
(trainData, testData, trainLabels, testLabels) = train_test_split(np.array(data),
	np.array(labels), test_size=0.25, random_state=42)

# initialize the model as a decision tree
model = DecisionTreeClassifier(random_state=84)
 
# check to see if a Random Forest should be used instead
if args["forest"] > 0:
	model = RandomForestClassifier(n_estimators=20, random_state=42)
 
# train the decision tree
print("[INFO] training model...")
model.fit(trainData, trainLabels)

# evaluate the classifier
print("[INFO] evaluating...")
predictions = model.predict(testData)
print(classification_report(testLabels, predictions))

# loop over a few random images
for i in np.random.randint(0, high=len(imagePaths), size=(10,)):
	# grab the image and classify it
	imagePath = imagePaths[i]
	filename = imagePath[imagePath.rfind("/") + 1:]
	image = cv2.imread(imagePath)
	features = describe(image)
	prediction = model.predict(features.reshape(1, -1))[0]
 
	# show the prediction
	print("[PREDICTION] {}: {}".format(filename, prediction))
	cv2.putText(image, prediction, (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 1.0, (0, 255, 0), 2)
	cv2.imshow("Image", image)
	cv2.waitKey(0)