detec.py

#! /usr/bin/env python
import os
import sys
import cv2
import numpy as np
from time import time
from datetime import datetime as dt
from tensorflow.keras.models import load_model
from tensorflow.keras.preprocessing.image import img_to_array
from tensorflow.keras.applications.mobilenet_v2 import preprocess_input

def detec(image):

    prototxt_path = os.path.sep.join([main_dir, 'Models/deploy.prototxt'])
    weights_path = os.path.sep.join([main_dir, 'Models/res10_300x300_ssd_iter_140000.caffemodel'])
    model = load_model(os.path.sep.join([main_dir, 'Models/model_detector.h5']))
    image = cv2.imread(image)
    (h, w) = image.shape[:2]

    net = cv2.dnn.readNetFromCaffe(prototxt_path, weights_path)  # Alterei de readNet para readNetFromCaffe

    blob = cv2.dnn.blobFromImage(cv2.resize(image, (300, 300)), 1.0, (300, 300), (104.0, 177.0, 123.0))
    net.setInput(blob)
    detections = net.forward()

    qtd_detec = 0
    labels = []
    for i in range(detections.shape[2]):
        confidence = detections[0, 0, i, 2]
        if confidence > 0.5:
            qtd_detec += 1

            box = detections[0, 0, i, 3:7] * np.array([w, h, w, h])
            (startX, startY, endX, endY) = box.astype("int")
            (startX, startY) = (max(0, startX), max(0, startY))
            (endX, endY) = (min(w - 1, endX), min(h - 1, endY))

            face = image[startY:endY, startX:endX]
            face = cv2.cvtColor(face, cv2.COLOR_BGR2RGB)
            face = cv2.resize(face, (224, 224))
            face = img_to_array(face)
            face = preprocess_input(face)
            face = np.expand_dims(face, axis=0)

            _, with_mask, without_mask = model.predict(face, batch_size=32)[0]

            if with_mask >= 0.85:
                color = (0, 128, 0)
                label = f"{with_mask*100:.1f}%: With mask"
            else:
                label = f"{(1-with_mask)*100:.1f}%: Without mask"
                color = (0, 0, 255)

            labels.append(label)

            aspect_ratio = float(image.shape[1]) / float(image.shape[0])
            cv2.putText(image, label, (startX, startY - 10),
                        cv2.FONT_HERSHEY_SIMPLEX, 1.5 * aspect_ratio, color, 4)
            cv2.rectangle(image, (startX, startY), (endX, endY), color, 10)

    return image, qtd_detec, labels

def loop_folder(folder_in):
    main_dir = os.getcwd()  # Adicionei essa linha para garantir que main_dir seja definido corretamente
    in_dir = f'{main_dir}/{folder_in}'
    out_dir = f'{in_dir}output'

    if not os.path.exists(out_dir):
        os.mkdir(out_dir)
        out_dir = out_dir

    f = open(f'{in_dir}log.txt', "w")
    f.write(dt.now().strftime('%Y-%m-%d %H:%M:%S\n'))

    t1 = time()
    for subdir, dirs, files in os.walk(in_dir):
        if subdir != out_dir:
            for file in files:
                
                if file.endswith((".jpg", ".png", ".jpeg")):
                    
                    img, qtd_detec, labels = detec(os.path.join(subdir, file))

                    os.chdir(out_dir)
                    filename = (f'N_{file}')
                    cv2.imwrite(filename, img)

                    f.write(f'\n{filename}: {qtd_detec} detections \n')
                    for l in labels:
                        f.write(f'{l}\n')
    t2 = time()
    print(f'I spent {t2 - t1:.2f}s on this joke')

def resize(image, window_height=720):
    aspect_ratio = float(image.shape[1]) / float(image.shape[0])
    window_width = window_height / aspect_ratio
    image = cv2.resize(image, (int(window_height), int(window_width)))
    return image

def main(pic_folder):
    main_dir = os.getcwd()  # Adicionei essa linha para garantir que main_dir seja definido corretamente
    
    if pic_folder.endswith("/"):
        
        loop_folder(pic_folder)

    else:
        
        image, *_ = detec(pic_folder)
        cv2.imshow('output', resize(image))
        cv2.waitKey(0)
        cv2.destroyAllWindows()

main_dir = os.getcwd()  # Adicionei essa linha para garantir que main_dir seja definido corretamente
main(sys.argv[1])