最近学习发现, 原来简单的人脸识别已经很成熟, 都不需要用到 AI 了, 直接有现成的库.
最开始是 win10+pycharm, 安装没搞定什么 cmake ,dlib,face_recogniton 库. 尝试在虚拟机 Ubuntu 中可以正常安装
- import numpy as np
- import argparse
- import cv2
- from PIL import Image
- import face_recognition
- # 参数导入指定图片
- ap=argparse.ArgumentParser()
- ap.add_argument("-i","--image",required=True,help="Enter path to the image")
- args=vars(ap.parse_args())
- image =cv2.imread(args["image"])
- #load the jpg file into a numpy
- #image=face_recognition.load_image_file("presidents.jpg")
- #find all the faces in the image using the default HOG-based model.
- # 自动识别图片中几个人脸, 每个人脸对应四个参数 上左下右 脸的边缘坐标点
- face_locations=face_recognition.face_locations(image)
- print("I found {} faces in this photograph,".format(len(face_locations)))
- for face_location in face_locations:
- #print the location of each face in this image
- top,right,bottom,left=face_location
- print("A face is located at pixel location Top:{},Left:{},Bottom:{},Right:{}".format(top,left,bottom,right))
- #access the actual face itself like this:
- face_image=image[top:bottom,left:right]
- pil_image=Image.fromarray(face_image)
- pil_image.show()
- ~
- ~
- ~
尝试用家人照片, 确实很快分割出来.
试试 dota1 的 coco 和 face void 人物造型
很快识别了船长的脸
虚空就识别不到. 初步看, 准确率还可以
来源: http://www.bubuko.com/infodetail-2910790.html