Extract Images from a Website in Java

Aspose.HTML for Java can load a web page, inspect its DOM, resolve relative resource URLs, and request image files. The examples on this page extract images referenced by <img src> and icons referenced by <link rel="icon">.

To extract images from a website in Java, load the page into an HTMLDocument, find the relevant <img> or <link> elements, read their src or href attributes with getAttribute(), resolve relative references against document.getBaseURI(), and download each resource through the document network service.

Extract Images from HTML Image Elements

Many images in an HTML document are referenced by the src attribute of an <img> element. The following example collects those attribute values in a Set, converts relative references to absolute URLs, requests each resource, and saves successful responses.

1. Create an HTMLDocument from the web page URL.
2. Call getElementsByTagName(“img”) to collect the <img> elements.
3. Read each element’s src attribute and collect distinct values.
4. Resolve every relative reference against the document base URI with the Url class.
5. Create and send a RequestMessage for each absolute URL.
6. For a successful ResponseMessage, derive an output name and write the response bytes to the destination directory.

// Extract images from website using Java

// Open a document you want to download images from
final HTMLDocument document = new HTMLDocument("https://docs.aspose.com/svg/net/drawing-basics/svg-shapes/");

// Collect all <img> elements
HTMLCollection images = document.getElementsByTagName("img");

// Create a distinct collection of relative image URLs
Iterator<Element> iterator = images.iterator();
java.util.Set<String> urls = new HashSet<>();
for (Element e : images) {
    urls.add(e.getAttribute("src"));
}

// Create absolute image URLs
java.util.List<Url> absUrls = urls.stream()
        .map(src -> new Url(src, document.getBaseURI()))
        .collect(Collectors.toList());

// foreach to while statements conversion
for (Url url : absUrls) {
    // Create an image request message
    final RequestMessage request = new RequestMessage(url);

    // Extract image
    final ResponseMessage response = document.getContext().getNetwork().send(request);

    // Check whether a response is successful
    if (response.isSuccess()) {
        String[] split = url.getPathname().split("/");
        String path = split[split.length - 1];

        // Save file to a local file system
        FileHelper.writeAllBytes(path, response.getContent().readAsByteArray());
    }
}

The example finds only values of <img src>. It does not parse srcset, <picture><source>, CSS background-image, or custom lazy-loading attributes such as data-src.

Always respect the website’s terms of use, copyright, and permissions when downloading or reusing image files.

Extract Icons from Website

Website icons can be referenced by <link> elements. The current example selects links whose rel attribute is exactly "icon", reads their href values, resolves the URLs, and downloads the resources.

1. Load the source web page into an HTMLDocument.
2. Collect its <link> elements.
3. Keep elements whose rel value equals "icon".
4. Read distinct href values and resolve them against document.getBaseURI().
5. Send a request for each resulting URL.
6. Save the bytes returned by successful responses.

// Download icons from website using Java

// Open a document you want to download icons from
final HTMLDocument document = new HTMLDocument("https://docs.aspose.com/html/net/message-handlers/");

// Collect all <link> elements
HTMLCollection links = document.getElementsByTagName("link");

// Leave only "icon" elements
java.util.Set<Element> icons = new HashSet<>();
for (Element link : links) {
    if ("icon".equals(link.getAttribute("rel"))) {
        icons.add(link);
    }
}

// Create a distinct collection of relative icon URLs
java.util.Set<String> urls = new HashSet<>();
for (Element icon : icons) {
    urls.add(icon.getAttribute("href"));
}

// Create absolute image URLs
java.util.List<Url> absUrls = urls.stream()
        .map(src -> new Url(src, document.getBaseURI()))
        .collect(Collectors.toList());

// foreach to while statements conversion
for (Url url : absUrls) {
    // Create a downloading request
    final RequestMessage request = new RequestMessage(url);

    // Extract icon
    final ResponseMessage response = document.getContext().getNetwork().send(request);

    // Check whether a response is successful
    if (response.isSuccess()) {
        String[] split = url.getPathname().split("/");
        String path = split[split.length - 1];

        // Save file to a local file system
        FileHelper.writeAllBytes(path, response.getContent().readAsByteArray());
    }
}

The exact comparison with "icon" does not cover every icon declaration. An HTML rel attribute can contain multiple tokens, and websites may use values such as shortcut icon or apple-touch-icon.

Image Sources Covered by the Examples

Image source	Covered?	Notes
<img src="image.png">	Yes	Relative and absolute src values are collected.
<link rel="icon" href="favicon.ico">	Yes	The icon example requires the exact rel="icon" value.
<img srcset="..."> or <picture>	No	Parse srcset and <source> elements separately.
CSS background-image	No	The image URL is stored in CSS rather than an HTML src attribute.
Lazy-loading attributes such as data-src	No	Attribute names are site-specific and require additional extraction logic.
Inline data: images	Not filtered	Decide whether to decode them or skip them before creating network requests.

Common Image Extraction Issues

Issue	Likely cause	Fix
The page URL is downloaded as though it were an image	An <img> has an empty src, which resolves to the document URL.	Remove null, empty, and whitespace-only attribute values before resolving URLs.
A data URI causes a request or file-name problem	An image is embedded directly in a data: URL.	Skip data URIs or decode them separately.
One downloaded image overwrites another	Different URLs have the same final path segment.	Generate unique output names instead of relying only on the URL file name.
Some visible images are missing	They are defined through srcset, CSS, JavaScript, or a custom lazy-loading attribute.	Inspect the loaded DOM and styles, then add extraction logic for the required source type.
A large image consumes substantial memory	readAsByteArray() buffers the complete response before saving.	Use an appropriate streaming workflow for large resources.

FAQ

Does the example extract every image visible on a website?

No. It extracts resources referenced by <img src> and exact <link rel="icon"> values. Responsive images, CSS backgrounds, data URIs, and site-specific lazy-loaded images require additional handling.

Can I extract image URLs from a local HTML file?

Yes. Load the local file into an HTMLDocument and use the same DOM queries. Relative references are resolved from the document base URI, and the referenced resources must remain accessible if the application also downloads them.

Related Aspose.HTML Articles

• Data Extraction in Java – explore other HTML and web data extraction workflows.
• Extract SVG from Website – extract inline SVG markup and external SVG resources.
• Save File from URL in Java – download a resource when its URL is already known.

Other Platforms

• Extract Images from Website in C# – extract and download website images with Aspose.HTML for .NET.

Try HTML Web Applications

Aspose.HTML provides free online HTML Web Applications for converting files, extracting web data, generating HTML, and analyzing pages without installing additional software.