使用Google Vision API调用特定图片URL返回空响应的解决方法

问题原因 #

juicer.io的图片链接存在防盗链机制：浏览器访问该链接时会携带Cookie、Referer等验证信息，而Google Vision API通过image_uri直接请求时，没有这些验证信息，服务器会拒绝返回图片内容，导致API返回空响应。

解决方案：先下载图片到本地再处理 #

绕过远程请求限制，先将图片下载到本地，再通过本地文件路径调用Vision API处理，具体步骤如下：

1. 安装requests库（用于下载图片）：

pip install requests

2. 修改代码，添加下载逻辑并读取本地图片：

from google.cloud import vision
import os
import sys
import requests

def download_image(url, save_path):
    # 模拟浏览器请求头，规避防盗链拦截
    headers = {
        "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/118.0.0.0 Safari/537.36",
        "Referer": "https://www.juicer.io/"
    }
    response = requests.get(url, headers=headers)
    response.raise_for_status()
    with open(save_path, "wb") as f:
        f.write(response.content)

if __name__ == '__main__':
    path = os.path.abspath(os.path.dirname(sys.argv[0])) 
    credsFN = os.path.join(path, "creds.json")
    os.environ['GOOGLE_APPLICATION_CREDENTIALS'] = credsFN

    # 目标juicer.io图片链接
    imageURL = "https://www.juicer.io/api/posts/459631754/images.jpg?external_id=CiE264cqZcw&s=199f57b090efba5ef332c3ff09d7b14554f84bfc"
    local_image_path = os.path.join(path, "temp_image.jpg")

    # 下载图片到本地
    try:
        download_image(imageURL, local_image_path)
    except Exception as e:
        print(f"下载图片失败: {e}")
        sys.exit(1)

    # 调用Vision API处理本地图片
    client = vision.ImageAnnotatorClient()
    with open(local_image_path, "rb") as image_file:
        content = image_file.read()
    image = vision.Image(content=content)
    
    features = [vision.Feature(type_=vision.Feature.Type.LABEL_DETECTION)]
    requ = vision.AnnotateImageRequest(image=image, features=features)
    resp = client.annotate_image(request=requ)
    
    # 输出标注结果
    if resp.label_annotations:
        for label in resp.label_annotations:
            print(f"{label.score:4.0%}: {label.description}")
    else:
        print("未检测到标注结果，请检查图片是否有效")
    
    # 可选：删除临时图片
    os.remove(local_image_path)

关键说明 #

• 伪装浏览器请求：通过User-Agent和Referer模拟浏览器访问，避免被juicer.io的防盗链机制拦截。
• 本地文件处理：Vision API读取本地图片时不受远程服务器限制，能稳定获取图片内容进行分析。

内容的提问来源于stack exchange，提问作者Rapid1898