Display images one by one with a next button

I have a streamlit code snippet below, where I iterate through a folder to display all the images there.

if st.button("Display images with PII"):
        out_folder = "output"
        for i in os.listdir(out_folder):
            if i.endswith(("jpg", "jpeg", "png")):
                image = Image.open(os.path.join(out_folder, i))
                st.image(image, caption=i)

However, I do not want to display all the images at one go, but rather one image at a time, preferably with a “Next” button to show the next image. Is this possible?

1 Like

I’ve been looking for something similar for days. How to show the images incrementally and not all at once. I have an application and sometimes need to show hundreds of images as the result of an action. But displaying those images takes a lot of time and impacts usability.
Is there anyway to be able to load the images incrementally, either by pushing a button (for instance, “show next 20 images”), or showing the images in multiple pages (for instance, something like this: First … 5 6 7 8 … Last)

Probably not the best approach but here is an idea using a session_state parameter as a counter:



import streamlit as st
import os

col1,col2 = st.columns(2)

if 'counter' not in st.session_state: 
    st.session_state.counter = 0

def showPhoto(photo):
    col1.write(f"Index as a session_state attribute: {st.session_state.counter}")
    ## Increments the counter to get next photo
    st.session_state.counter += 1
    if st.session_state.counter >= len(pathsImages):
        st.session_state.counter = 0

# Get list of images in folder
folderWithImages = r"images"
pathsImages = [os.path.join(folderWithImages,f) for f in os.listdir(folderWithImages)]

col1.subheader("List of images in folder")

# Select photo a send it to button
photo = pathsImages[st.session_state.counter]
show_btn = col1.button("Show next pic ⏭️",on_click=showPhoto,args=([photo]))
1 Like

Thanks @edsaac

Using your example I was able to build the functionality that I was looking for. I’m still curious to know whether there is a recommended approach which is supported natively by streamlit, or if they’re working on that.