實(shí)時(shí)監(jiān)控圖像中的人臉識(shí)別：理解人臉識(shí)別技術(shù)指南
在當(dāng)今的數(shù)字時(shí)代，人臉識(shí)別技術(shù)已經(jīng)成為一個(gè)關(guān)鍵技術(shù)，它正在從安全到個(gè)性化體驗(yàn)等多個(gè)領(lǐng)域進(jìn)行革新。從門(mén)禁控制到考勤系統(tǒng)，再到走失兒童的檢測(cè)，人臉識(shí)別的應(yīng)用非常廣泛，涵蓋了安全、教育和公共安全領(lǐng)域。然而，在不同的條件下準(zhǔn)確識(shí)別人臉面臨著一系列獨(dú)特的挑戰(zhàn)。
在本指南中，我們將深入探討人臉識(shí)別的復(fù)雜性，探索使用Python和先進(jìn)的機(jī)器學(xué)習(xí)工具來(lái)確保強(qiáng)大的識(shí)別和驗(yàn)證的技術(shù)。
理解人臉識(shí)別
人臉識(shí)別技術(shù)利用先進(jìn)的算法分析和比較從圖像或視頻片段中提取的面部特征，從而實(shí)現(xiàn)個(gè)人的識(shí)別和驗(yàn)證。提供的Python代碼展示了人臉識(shí)別的實(shí)現(xiàn)。讓我們探索代碼的主要組成部分，以深入了解該過(guò)程。
1. 導(dǎo)入庫(kù)
在本節(jié)中，導(dǎo)入了執(zhí)行各種任務(wù)所需的庫(kù)，如面部檢測(cè)、圖像處理、數(shù)據(jù)增強(qiáng)和人臉識(shí)別。
import face_recognition
import cv2
import numpy as np
from deepface import DeepFace
from utils import apply_blur, generate_unique_random_numbers, find_cosine_distance_helper
from utils import apply_resize
from utils import augment_data, face_distance
import os
from PIL import Image
from mtcnn.mtcnn import MTCNN
import random2.數(shù)據(jù)集準(zhǔn)備
數(shù)據(jù)集準(zhǔn)備階段涉及遍歷包含已知個(gè)人圖像的目錄(known_people_dir)。對(duì)于每個(gè)人，代碼在train_dataset目錄中創(chuàng)建一個(gè)輸出目錄。然后加載每張圖像，使用MTCNN(多任務(wù)級(jí)聯(lián)卷積網(wǎng)絡(luò))模型檢測(cè)面部，裁剪檢測(cè)到的面部區(qū)域，并將其保存在相應(yīng)的輸出目錄(known_people_train_dir)中。此外，還應(yīng)用了模糊、調(diào)整大小和應(yīng)用隨機(jī)變換等數(shù)據(jù)增強(qiáng)技術(shù)，以增加數(shù)據(jù)集的多樣性并增強(qiáng)人臉識(shí)別系統(tǒng)的魯棒性。
for person_name in os.listdir(known_people_dir):
person_dir = os.path.join(known_people_dir, person_name)
if os.path.isdir(person_dir):
output_person_dir = os.path.join("train_dataset", person_name)
os.makedirs(output_person_dir, exist_ok=True)
for filename in os.listdir(person_dir):
image_path = os.path.join(person_dir, filename)
image = cv2.imread(image_path)
faces = mtcnn.detect_faces(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))
if faces:
for face in faces:
x, y, w, h = face['box']
left = max(x, 0)
top = max(y, 0)
right = min(x + w, image.shape[1])
bottom = min(y + h, image.shape[0])
if right > left and bottom > top:
output_face_path = os.path.join(output_person_dir, f"{filename}.jpg")
cv2.imwrite(output_face_path, image)
# Apply data augmentation
apply_blur(output_face_path, output_folder=output_person_dir)
apply_resize(output_face_path, output_folder=output_person_dir)
augment_data(output_face_path, output_folder=output_person_dir, face_coordinates=(left, top, right, bottom), prefix=filename)
訓(xùn)練數(shù)據(jù)集
3. 檢測(cè)面部
使用MTCNN(多任務(wù)級(jí)聯(lián)卷積網(wǎng)絡(luò))模型進(jìn)行面部檢測(cè)，該模型能夠檢測(cè)圖像中的面部。然后使用檢測(cè)到的面部進(jìn)行進(jìn)一步處理。
# Detect faces in the image using MTCNN
faces = mtcnn.detect_faces(rgb_image)4.提取邊界框坐標(biāo)：
# Get the bounding box coordinates of the face
x, y, w, h = face['box']
# Ensure that the bounding box coordinates are valid
left = max(x, 0)
top = max(y, 0)
right = min(x + w, image.shape[1])
bottom = min(y + h, image.shape[0])
圖片
5. 圖像增強(qiáng)
對(duì)提取的面部圖像應(yīng)用模糊、調(diào)整大小和隨機(jī)變換等數(shù)據(jù)增強(qiáng)技術(shù)，以增強(qiáng)數(shù)據(jù)集的多樣性。
(1) 隨機(jī)變換
def augment_data(original_image_path, output_folder, face_coordinates, num_augmented_images=3, should_add_jitter=True, prefix=""):
# Load the original image
original_image = Image.open(original_image_path)
# Convert face image to grayscale
face_image_gray = original_image.convert('L')
# Define torchvision transforms for data augmentation
data_transforms = transforms.Compose([
transforms.RandomHorizontalFlip(),
transforms.RandomRotation(degrees=15),
transforms.ColorJitter(brightness=0.2, contrast=0.2, saturation=0.2, hue=0.1),
transforms.ToTensor(),
])
# Apply data augmentation and save augmented images
for i in range(num_augmented_images):
# Apply different transformations to each augmented image
transformed_image = data_transforms(face_image_gray)
augmented_image_path = os.path.join(output_folder, f"{prefix}augmented_{i + 1}.jpg")
transforms.ToPILImage()(transformed_image).save(augmented_image_path)
print(f"Augmented image {i + 1} saved to {augmented_image_path}")
(2) 調(diào)整大小和模糊
def apply_blur(image_path, output_folder, kernel_size=(7, 7)):
# Load the image
image = cv2.imread(image_path)
# Apply Gaussian blur
blurred_image = cv2.GaussianBlur(image, kernel_size, 0)
# Save the blurred image
filename = os.path.basename(image_path)
output_path = os.path.join(output_folder, f"blurred_{filename}")
cv2.imwrite(output_path, blurred_image)
print(f"Blurred image saved to {output_path}")
def apply_resize(image_path, output_folder, target_size=(256, 256)):
# Load the image
image = cv2.imread(image_path)
# Resize the image
resized_image = cv2.resize(image, target_size)
# Save the resized image
filename = os.path.basename(image_path)
output_path = os.path.join(output_folder, f"resized_{filename}")
cv2.imwrite(output_path, resized_image)
print(f"Resized image saved to {output_path}")
6.在訓(xùn)練目錄中存儲(chǔ)增強(qiáng)圖像
處理后的圖像，包括裁剪的面部、模糊的面部和增強(qiáng)的圖像，存儲(chǔ)在訓(xùn)練目錄(train_dataset)中。這種目錄結(jié)構(gòu)便于訪問(wèn)訓(xùn)練數(shù)據(jù)，以構(gòu)建人臉識(shí)別模型。
# Save the face image
cv2.imwrite(output_face_path, image)# Apply data augmentation on the face image
apply_blur(output_face_path, output_folder=output_person_dir)
apply_resize(output_face_path, output_folder=output_person_dir)
augment_data(output_face_path, output_folder=output_person_dir,
face_coordinates=(left, top, right, bottom),prefix=filename)7.從訓(xùn)練數(shù)據(jù)集中編碼已知面部
代碼遍歷我們指定的訓(xùn)練數(shù)據(jù)集目錄中的目錄，稱為known_people_train_dir。在每個(gè)代表特定個(gè)人的目錄中，它處理每個(gè)圖像文件。代碼驗(yàn)證每個(gè)圖像文件的有效性，加載它，并使用高級(jí)技術(shù)提取面部特征。這些特征被編碼成數(shù)值向量，稱為面部編碼，使用DeepFace.represent函數(shù)。這些編碼以及相應(yīng)的人名然后被添加到列表中以進(jìn)行進(jìn)一步處理。
通過(guò)將增強(qiáng)數(shù)據(jù)與原始圖像結(jié)合，我們的模型訓(xùn)練數(shù)據(jù)集變得更加豐富和多樣化，從而在不同條件和環(huán)境中提高了人臉識(shí)別的準(zhǔn)確性和魯棒性。
for person_name in os.listdir(known_people_train_dir):
person_dir = os.path.join(known_people_train_dir, person_name)
# Check if it's a directory
if os.path.isdir(person_dir):
# Iterate over each file in the person's directory
for filename in os.listdir(person_dir):
image_path = os.path.join(person_dir, filename)
print(image_path)
# Check if the file is a valid image file
try:
with Image.open(image_path) as img:
img.verify() # Attempt to open and verify the image file
# Load the image file
person_image = face_recognition.load_image_file(image_path)
# Encode the face in the image-
face_encoding = DeepFace.represent(person_image,model_name="Dlib",detector_backend="mtcnn", enforce_detection=False)
# Append the face encoding and name to the respective lists
known_face_encodings.append(np.array(face_encoding[0]['embedding']))
known_face_names.append(person_name)
except (IOError, SyntaxError,IndexError):
# Ignore any files that are not valid image files
continue8.人臉識(shí)別循環(huán)
在人臉識(shí)別循環(huán)中，程序不斷從網(wǎng)絡(luò)攝像頭捕獲幀，確保實(shí)時(shí)人臉識(shí)別。為了優(yōu)化處理速度，每個(gè)幀都被調(diào)整大小，減少了計(jì)算負(fù)載而不影響準(zhǔn)確性。使用MTCNN面部檢測(cè)模型，程序在幀內(nèi)識(shí)別面部，對(duì)其特征進(jìn)行編碼以進(jìn)行比較。
# Continuous capture of frames from the webcam
while True:
ret, frame = video_capture.read()
# Resize each frame for optimized processing speed
small_frame = cv2.resize(frame, (0, 0), fx=0.25, fy=0.25)
# Using MTCNN for face detection
rgb_small_frame = small_frame[:, :, ::-1]
result1 = DeepFace.represent(rgb_small_frame, model_name="Dlib", detector_backend="mtcnn", enforce_detection=False)
# Encoding features of detected faces for comparison
face_locations = [(res['facial_area']['y'], res['facial_area']['x'] + res['facial_area']['w'], res['facial_area']['y'] + res['facial_area']['h'], res['facial_area']['x']) for res in result1]
face_encodings = [res['embedding'] for res in result1]通過(guò)計(jì)算檢測(cè)到的面部和訓(xùn)練數(shù)據(jù)集中已知面部之間的余弦距離，程序確定潛在的匹配項(xiàng)。
# Calculating cosine distances between detected faces and known faces
for f_encoding in face_encodings:
face_distances = find_cosine_distance_helper(known_face_encodings, f_encoding)
best_match_index = np.argmin(face_distances)
if face_distances[best_match_index] <= 0.07:
name = known_face_names[best_match_index]
else:
name = "Unknown"
face_names.append(name)9.顯示結(jié)果
檢測(cè)到的面部顯示在視頻流中，包括相應(yīng)的名稱(如果識(shí)別出來(lái)，否則為“未知”)。在面部周圍繪制矩形框，并在每個(gè)面部下方添加標(biāo)簽，以便于識(shí)別。
# Draw a bounding box around the face
cv2.rectangle(frame, (left, top), (right, bottom), (0, 0, 255), 2)
# Draw a label with a name below the face
cv2.putText(frame, text, (left + 6, bottom - 6), font, 1.0, (255, 255, 255), 1)
# Display the resulting image
cv2.imshow('Video', frame)輸出
人臉識(shí)別系統(tǒng)使用網(wǎng)絡(luò)攝像頭在實(shí)時(shí)面部檢測(cè)和識(shí)別任務(wù)中取得了顯著的性能。它能夠準(zhǔn)確識(shí)別已知個(gè)人，并以良好的精度進(jìn)行標(biāo)記，并將未知面部適當(dāng)?shù)貥?biāo)記為“未知”。系統(tǒng)以高置信度運(yùn)行，提高了其可靠性和可用性。在Salman的圖像上訓(xùn)練的模型準(zhǔn)確地識(shí)別了他在監(jiān)控錄像中的面部。
當(dāng)遇到不在數(shù)據(jù)集中的Amitabh的圖像時(shí)，它被適當(dāng)?shù)貥?biāo)記為“未知”。
