prediction.py

## Imports for making predictions
import os
import numpy as np
from PIL import Image
from keras.applications.imagenet_utils import preprocess_input
from keras.models import load_model
from keras.preprocessing import image
import cv2
from create_data import show_images
import pickle
from edge_detection import final_spot_dict
cwd = os.getcwd()
top_model_weights_path = 'car1.h5'

class_dictionary = {}
class_dictionary[0] = 'empty'
class_dictionary[1] = 'occupied'

from PIL import Image
model = load_model(top_model_weights_path)

from create_data import test_images
# pickel load

def make_prediction(image):
    #Rescale image
    img = image/255.

    #Convert to a 4D tensor
    image = np.expand_dims(img, axis=0)
    #print(image.shape)

    # make predictions on the preloaded model
    class_predicted = model.predict(image)
    inID = np.argmax(class_predicted[0])
    label = class_dictionary[inID]
    return label

empty_slots = []
def predict_on_image(image, spot_dict = final_spot_dict, make_copy=True, color = [0, 255, 0], alpha=0.5):
    if make_copy:
        new_image = np.copy(image)
        overlay = np.copy(image)
    cnt_empty = 0
    all_spots = 0
    for spot in spot_dict:
        all_spots += 1
        (x1, y1, x2, y2) = spot
        (x1, y1, x2, y2) = (int(x1), int(y1), int(x2), int(y2))
        #crop this image
        spot_img = image[y1:y2, x1:x2]
        spot_img = cv2.resize(spot_img, (48, 48)) 
        
        label = make_prediction(spot_img)
#         print(label)
        if label == 'empty':
            empty_slots.append(spot)
            cv2.rectangle(overlay, (int(x1),int(y1)), (int(x2),int(y2)), color, -1)
            cnt_empty += 1
            
    cv2.addWeighted(overlay, alpha, new_image, 1 - alpha, 0, new_image)
            
    cv2.putText(new_image, "Available: %d spots" %cnt_empty, (30, 95),
    cv2.FONT_HERSHEY_SIMPLEX,
    0.7, (255, 255, 255), 2)
    
    cv2.putText(new_image, "Total: %d spots" %all_spots, (30, 125),
    cv2.FONT_HERSHEY_SIMPLEX,
    0.7, (255, 255, 255), 2)
    save = False
    
    if save:
        filename = 'with_marking.jpg'
        cv2.imwrite(filename, new_image)
    
    return new_image


def sample_func():
    predicted_images = list(map(predict_on_image, test_images))
    show_images(predicted_images)
    with open('free_spots.pickle', 'wb') as handle:
        pickle.dump(empty_slots, handle, protocol=pickle.HIGHEST_PROTOCOL)



sample_func()