65行python代码实现换脸

素颜马尾好姑娘i 2024-04-18 18:02 110阅读 0赞

下面来介绍一种对照片的简单换脸方法。

**本文仅供学习，请勿用于侵权行为。**

本篇介绍的换脸方法需要借助Face++，关于Face++的API，大家可自行查看说明文档，都比较简单。

文档地址：https://console.faceplusplus.com.cn/documents/20813963

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQ0NjIxNTEw_size_16_color_FFFFFF_t_70]

import requests
    import simplejson
    import json
    import base64
     
    #Face++网址：[url]https://console.faceplusplus.com.cn/dashboard[/url]
    #第一步，获取人脸关键点
    def find_face(imgpath):
        print("正在查找……")
        http_url = 'https://api-cn.faceplusplus.com/facepp/v3/detect'
        data = {"api_key": '自己申请',
                "api_secret": '自己申请TLJC',
                "image_url": imgpath, "return_landmark": 1}
        files = {"image_file": open(imgpath, "rb")}
        response = requests.post(http_url, data=data, files=files)
        req_con = response.content.decode('utf-8')
        req_dict = json.JSONDecoder().decode(req_con)
        this_json = simplejson.dumps(req_dict)
     
        this_json2 = simplejson.loads(this_json)
        print(this_json2)
        faces = this_json2['faces']
        list0 = faces[0]
        rectangle = list0['face_rectangle']
        #print(rectangle)
        return rectangle
     
    #第二步，换脸，其中图片的大小应不超过2M
    # number表示换脸的相似度
    def merge_face(image_url1, image_url2, image_url, number):
     
        ff1 = find_face(image_url1)
        ff2 = find_face(image_url2)
     
        rectangle1 = str(str(ff1['top']) + "," + str(ff1['left']) + "," + str(ff1['width']) + "," + str(ff1['height']))
        rectangle2 = str(ff2['top']) + "," + str(ff2['left']) + "," + str(ff2['width']) + "," + str(ff2['height'])
        print(rectangle2)
        url_add = "https://api-cn.faceplusplus.com/imagepp/v1/mergeface"
        f1 = open(image_url1, 'rb')
        f1_64 = base64.b64encode(f1.read())
        f1.close()
        f2 = open(image_url2, 'rb')
        f2_64 = base64.b64encode(f2.read())
        f2.close()
     
        data = {"api_key": '自己申请',
                "api_secret": '自己申请TLJC',
                "template_base64": f1_64, "template_rectangle": rectangle1,
                "merge_base64": f2_64, "merge_rectangle": rectangle2, "merge_rate": number}
     
        response = requests.post(url_add, data=data)
        req_con1 = response.content.decode('utf-8')
        #print(req_con1)
        req_dict = json.JSONDecoder().decode(req_con1)
        result = req_dict['result']
        imgdata = base64.b64decode(result)
        file = open(image_url, 'wb')
        file.write(imgdata)
        file.close()
     
    image1 = r"E:\test\img01.jpg"
    image2 = r"E:\test\img02.jpg"
    image = r"E:\test\imgh.jpg"
     
    merge_face(image2, image1, image, 90)

参考：https://www.52pojie.cn/thread-1016974-1-2.html

[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQ0NjIxNTEw_size_16_color_FFFFFF_t_70]: https://image.dandelioncloud.cn/pgy_files/images/2024/04/18/e084518ccf5e4e1c8b52288cf80854ef.png