Streamlit上传图片报错：'UploadedFile'对象无'height'属性问题求助

Hey there! Let's sort out that AttributeError: 'UploadedFile' object has no attribute 'height' issue you're facing. The problem boils down to a simple misunderstanding of what st.file_uploader returns:

Why the Error Happens #

Streamlit's st.file_uploader gives you an UploadedFile object, which is essentially a file-like stream of your uploaded image—it's not a ready-to-use image object with properties like height or width. That's why trying to access img.height or convert it directly with asarray(img) won't work.

The Fix: Convert to a PIL Image First #

You need to first convert the uploaded file stream into a proper image object using PIL (Pillow), which will give you access to the image's metadata and let you convert it to a numpy array correctly. Here's how to adjust your code:

import streamlit as st
from PIL import Image
import numpy as np
from numpy import asarray

st.title("Where's Waldo?")
uploaded_picture = st.file_uploader("Upload a picture of the following game: Where's Waldo?", type=['png', 'jpg'])

# Always check if a file was uploaded before processing
if uploaded_picture is not None:
    # Step 1: Convert the UploadedFile to a PIL Image
    img = Image.open(uploaded_picture)
    
    # Step 2: Convert the PIL Image to a numpy array (for your recognition logic)
    img_array = asarray(img)
    
    # Step 3: Get image dimensions (you can use either PIL's attributes or the array's shape)
    # Option 1: From PIL Image
    # img_height, img_width = img.height, img.width
    # Option 2: From numpy array (more useful if you're working with the array later)
    img_height, img_width = img_array.shape[:2]
    
    # Now generate your output_array with the correct dimensions
    output_array = np.zeros((int(img_height / 64), int(img_width / 64)))
    
    # Optional: Display the uploaded image to confirm
    st.image(img, caption="Uploaded Waldo Image", use_column_width=True)
    
    # Continue with your Waldo detection logic here...

Key Notes #

• Install Pillow if you haven't: Run pip install pillow in your environment to use the PIL.Image module.
• Check for None: Adding if uploaded_picture is not None: prevents errors when the user hasn't uploaded an image yet.
• Dimension Options: You can get height/width from either the PIL Image (img.height) or the numpy array (img_array.shape[0] for height, shape[1] for width)—use whichever fits your workflow better.

This should resolve the AttributeError and let you process the uploaded image just like you did with local file paths!

内容的提问来源于stack exchange，提问作者Dolev