import numpy as np import argparse from GpuConfig import GpuMemoryAllocate from keras.callbacks import ModelCheckpoint ap = argparse.ArgumentParser() ap.add_argument("-d", "--dataset", required=True, help = "path to the dataset") ap.add_argument("-w", "--weights", required=True, help="path to best model weights file") args = vars(ap.parse_args()) epoch = 50 print("[INFO] loading images...") imagePaths = list(paths.list_images(args["dataset"])) sp = SimplePreprocessor(32, 32) iap = ImageArrayPreprocessor() sd1 = SimpleDatasetLoader(preprocessors=[sp, iap]) (data, labels) = sd1.load(imagePaths, Verbose=500) (trainX, testX, trainY, testY) = train_test_split(data, labels, test_size=0.25, random_state=42) trainX = trainX.astype("float") / 255.0 testX = testX.astype("float") / 255.0 labelNames = ["cat", "dog"] labelsOneHot_train = [] labelsOneHot_test = [] for label in trainY: labelOneHot = [1, 0] if label == 'cats' else [0, 1]
import cv2 from ImageTools import SimplePreprocessor from keras.preprocessing.image import img_to_array image = cv2.imread('/home/a/animals/training_set/training_set/cats/cat.2.jpg') image_array = img_to_array(image) cv2.imshow('image', image) cv2.waitKey() print(image_array.shape) print(image.shape) p = SimplePreprocessor(32, 32) image = p.preprocess(image) cv2.imshow('image', image) cv2.waitKey()