人工智能opencv 使用深度学习进行人脸性别的检测

使用OpenCV结合深度学习进行人脸性别检测的流程可分为人脸检测、性别模型推理、结果可视化三个核心步骤，具体实现方法如下：一、核心步骤解析人脸检测阶段使用预训练的Caffe模型（如res10_300x300_ssd_iter_140000.caffemodel）定位图像中的人脸区域。通过cv2.dnn.blobFromImage对输入图像进行预处理（尺寸缩放至300×300、RGB通道均值减除），再通过forward()获取检测框坐标。图1 人脸检测流程示意图性别检测阶段对每个检测到的人脸区域进行二次预处理（尺寸调整为227×227，使用BGR通道均值[78.426, 87.769, 114.896]），输入性别分类模型（如gender_net.caffemodel）获取概率向量，通过argmax()确定性别类别。结果可视化在原始图像上绘制检测框，并叠加性别标签及置信度文本（如Female: 98.32%），最终通过cv2.imshow()展示结果。二、代码实现细节模型初始化import cv2import numpy as np# 加载人脸检测模型faceNet = cv2.dnn.readNet("model/deploy.prototxt", "model/res10_300x300_ssd_iter_140000.caffemodel")# 加载性别检测模型genderNet = cv2.dnn.readNet("model/deploy_gender.prototxt", "model/gender_net.caffemodel")genderList = [Male, Female]图像预处理与推理image = cv2.imread("image/img.jpg")(h, w) = image.shape[:2]# 人脸检测预处理blob = cv2.dnn.blobFromImage(image, 1.0, (300, 300), (104.0, 177.0, 123.0))faceNet.setInput(blob)detections = faceNet.forward()性别分类与结果标注for i in range(detections.shape[2]): confidence = detections[0, 0, i, 2] if confidence 0.5: # 解析检测框坐标 box = detections[0, 0, i, 3:7] * np.array([w, h, w, h]) (startX, startY, endX, endY) = box.astype("int") face = image[startY:endY, startX:endX] # 性别检测预处理 faceBlob = cv2.dnn.blobFromImage(face, 1.0, (227, 227), (78.426, 87.769, 114.896), swapRB=False) genderNet.setInput(faceBlob) preds = genderNet.forward() # 获取分类结果 genderIdx = preds[0].argmax() gender = genderList[genderIdx] confidence = preds[0][genderIdx] * 100 # 绘制结果 text = f"{gender}: {confidence:.2f}%" y = startY - 10 if startY 10 else startY + 10 cv2.rectangle(image, (startX, startY), (endX, endY), (0, 0, 255), 2) cv2.putText(image, text, (startX, y), cv2.FONT_HERSHEY_SIMPLEX, 0.45, (0, 0, 255), 2)结果展示cv2.imshow("Output", image)cv2.waitKey(0)cv2.destroyAllWindows()三、关键参数说明blobFromImage参数解析scalefactor=1.0：图像缩放系数（实际值为1/σ）size