如何通过浏览器在服务器上运行Python版Google Vision云端应用

Hey there! Since you're new to Python and trying to get Google Vision's landmark detection up and running on your server (http://18.217.137.107), let's break this down into simple, step-by-step actions you can follow. I'll cover everything from initial setup to testing the app.

First, you need a service account key to authenticate your server with Google Vision API. Here's how to get it:

• Head to the Google Cloud Console, select your project (or create one if you haven't), then navigate to IAM & Admin > Service Accounts.
• Create a new service account, assign it the Cloud Vision API User role, then download the JSON key file.

Next, get this key onto your server:

• Use scp from your local machine to upload the file:

  scp /path/to/your-service-key.json your-server-username@18.217.137.107:/path/you/want/on/server/
  

• Log into your server via ssh, then set an environment variable so the Google SDK can find the key:

  export GOOGLE_APPLICATION_CREDENTIALS="/path/you/want/on/server/your-service-key.json"
  

  (To make this permanent, you can add this line to your ~/.bashrc or ~/.profile file and run source ~/.bashrc.)

Make sure your server has Python and the necessary packages:

1. If Python/pip isn't installed (Ubuntu example):

   sudo apt update && sudo apt install python3 python3-pip -y
   

2. Install the Google Cloud Vision library:

   pip3 install google-cloud-vision
   

Let's create a basic Python script to test landmark detection. Create a file named landmark_detector.py on your server with this code:

import io
from google.cloud import vision_v1

def detect_landmarks(image_path):
    """Detects landmarks in a local image file."""
    # Initialize the Vision client
    client = vision_v1.ImageAnnotatorClient()

    # Read the image content
    with io.open(image_path, 'rb') as image_file:
        content = image_file.read()

    image = vision_v1.Image(content=content)

    # Send the image to Google Vision for analysis
    response = client.landmark_detection(image=image)
    landmarks = response.landmark_annotations

    # Print the results
    print("Detected Landmarks:")
    for landmark in landmarks:
        print(f"- {landmark.description}")
        for location in landmark.locations:
            lat_lng = location.lat_lng
            print(f"  Coordinates: Lat {lat_lng.latitude}, Lng {lat_lng.longitude}")

    # Handle errors
    if response.error.message:
        raise Exception(f"Vision API Error: {response.error.message}")

# Test with a local image
if __name__ == "__main__":
    # Replace 'test_landmark.jpg' with your image file path
    detect_landmarks("test_landmark.jpg")

Now, upload a test image (like a photo of the Eiffel Tower or Mount Everest) to the same directory as the script, naming it test_landmark.jpg.

On your server, navigate to the directory with the script and image, then run:

python3 landmark_detector.py

If everything works, you should see the landmark name and coordinates printed in your terminal.

If you want to access this functionality via your server's URL (http://18.217.137.107), you can wrap the code in a simple Flask web app:

1. Install Flask:

   pip3 install flask
   

2. Create a file named app.py with this code:

import io
from flask import Flask, request, jsonify
from google.cloud import vision_v1

app = Flask(name)
client = vision_v1.ImageAnnotatorClient()

@app.route('/detect-landmarks', methods=['POST'])
def detect_landmarks_api():
if 'image' not in request.files:
return jsonify({"error": "No image file provided"}), 400

image_file = request.files['image']
content = image_file.read()
image = vision_v1.Image(content=content)

response = client.landmark_detection(image=image)
landmarks = response.landmark_annotations

result = []
for landmark in landmarks:
    landmark_info = {
        "name": landmark.description,
        "coordinates": []
    }
    for loc in landmark.locations:
        lat_lng = loc.lat_lng
        landmark_info["coordinates"].append({
            "latitude": lat_lng.latitude,
            "longitude": lat_lng.longitude
        })
    result.append(landmark_info)

if response.error.message:
    return jsonify({"error": response.error.message}), 500

return jsonify({"detected_landmarks": result})

if name == "main":
# Run the server on all interfaces so it's accessible via your server IP
app.run(host='0.0.0.0', port=8080)

3. Run the app:
```bash
python3 app.py

4. Test it from your local machine with curl:

   curl -X POST -F "image=@/path/to/your/test/image.jpg" http://18.217.137.107:8080/detect-landmarks
   

   You'll get a JSON response with the detected landmarks.

• Authentication Errors: Double-check that the GOOGLE_APPLICATION_CREDENTIALS path is correct, and the key file has proper permissions (chmod 600 your-service-key.json to restrict access).
• Package Installation Issues: Upgrade pip first with pip3 install --upgrade pip if you run into errors installing google-cloud-vision.
• Server Access: Ensure your server's firewall allows outgoing connections to Google's API endpoints, and incoming connections on port 8080 (if you're running the web app).

内容的提问来源于stack exchange，提问作者user9394885