check_network.py

import cv2
import numpy as np
from network import Network


num_pixels = 39900
num_aircraft = 100

image_network = Network(num_pixels, num_aircraft, 0.05, hid_out=np.loadtxt("hid_out.txt"), in_hid=np.loadtxt("in_hid.txt"))
images = []  # IMAGES HERE! PLEASE IN THE SAME ORDER AS IN THE GOOGLE DOC!!
# all image inputs must be arrays with each element representing % in a single pixel (grayscale)

for i in range(1, num_aircraft + 1):
    name = str(i)
    if i < 10:
        name = "0" + name
    img = cv2.imread("pics_processed/" + name + ".jpg", 0)
    pixels = []
    for j in range(img.shape[0]):  # traverses through height of the image
        for k in range(img.shape[1]):  # traverses through width of the image
            pixels.append(img[i][j])
    pixels = np.transpose(pixels)
    images.append(pixels)

aircraft = [0] * 10 + [1] * 7 + [2] * 9 + [3] * 6 + [4] * 4 + [5] + [4] + [6] + [4] + [2] + [4] * 2 + [7] + [5] * 2 + [
    6] + [8] * 2 + [9] * 3 + [10] * 3 + [11] * 2 + [12] * 3 + [13] + [14] * 3 + [8] * 2 + [15] * 5 + [16] * 3 + [
    17] * 3 + [18] * 2 + [19] * 2 + [20] * 4 + [21] * 3 + [22] * 3 + [23] * 4 + [24] * 4 + [25]  # actual aircraft


planes = {
    '0': "B737",
    '1': "A320",
    '2': "B777",
    '3': "A380",
    '4': "B747",
    '5': "An-255",
    '6': "Airbus Beluga",
    '7': "B767",
    '8': "CRJ",
    '9': "E175",
    '10': "Cessna Skyhawk",
    '11': "Cessna Citation",
    '12': "Cessna Skymaster",
    '13': "Piper M600",
    '14': "A220",
    '15': "A319",
    '16': "Bell 407",
    '17': "Airbus H120",
    '18': "Embraer Phemom",
    '19': "HA420",
    '20': "DC10",
    '21': "Aquila A210",
    '22': "An-148",
    '23': "ATR 42",
    '24': "Dash-8",
    '25': "Beechcraft B200GT",
}

for index, image in enumerate(images):
    hid, out = image_network.predict(image)
    this_aircraft = np.argmax(out)
    plane = planes[str(this_aircraft)]

    print(index, "Actually", planes[str(aircraft[index])], "Predicted", plane, sep="\t\t\t")