Given an IMDB movie id, how do I programmatically get its poster image?

Question

movie id tt0438097 can be found at http://www.imdb.com/title/tt0438097/

What's the url for its poster image?

Katzman

why not use a different source for the posters like themoviedb.org? – Atelectasis

any full source code sample ? – Waterage

Answer 1

As I'm sure you know, the actual url for that image is

http://ia.media-imdb.com/images/M/MV5BMTI0MDcxMzE3OF5BMl5BanBnXkFtZTcwODc3OTYzMQ@@._V1._SX100_SY133_.jpg

You're going to be hard pressed to figure out how it's generated though and they don't seem to have a publicly available API.

Screenscraping is probably your best bet.

The picture seems to generally be inside a div with class=photo and the name of the a tag is poster.

The image itself is just inside the a tag.

Overestimate

Hard pressed? Doubt it. That URL is clearly Base64 encoded. The last part X-Y is the pictures size. Just need to dig around, I'm sure there are enough clues to decode it. – Insurgence

Yup! I was right. "MV5BMTI0MDcxMzE3OF5BMl5BanBnXkFtZTcwODc3OTYz" decodes to "1^A1240713178^A2^Ajpg^Ame70877963" So there is another piece to your puzzle. – Insurgence

Answer 2

Check out http://www.imdbapi.com/, It returns Poster url in string.

For example, check http://www.imdbapi.com/?i=&t=inception and you'll get the poster address: Poster":"http://ia.media-imdb.com/images/M/MV5BMjAxMzY3NjcxNF5BMl5BanBnXkFtZTcwNTI5OTM0Mw@@._V1._SX320.jpg"

Update: Seems like the site owner had some arguments with IMDB legal staff. As mentioned in the original site, new site's address is http://www.omdbapi.com/

Cristen

This site doesn't look like it has established itself well enough to be considered reliable yet, but I look forward to seeing how it develops. Thanks for the link! – Passepartout

Notice that you cannot rely or control the poster size/cropping dimensions of the posters. This will be a pain in the ass if you need same size for thumbnail grid. For instance, this title will give out horizontal image despite the main image of that short movie is cropped differently on the page. It would really help to know the sizing and cropping logic of the poster, embedded in the image url. – Epsomite

but those links cannot be used anyway : #28677108 – Pagoda

Answer 3

The best solution is to use tmdb.org :

1. use your imdbid in this api url after find/:

   https://api.themoviedb.org/3/find/tt0111161?api_key=__YOURAPIKEY__&external_source=imdb_id

2. Retrieve the json response and select the poster_path attribute:

   e.g. "poster_path":"/9O7gLzmreU0nGkIB6K3BsJbzvNv.jpg"

3. Prepend this path with "http://image.tmdb.org/t/p/w150", and you will have the poster URL that you can use in an img tag :-)

---

omdbapi works, but I found out you cannot really use these images (because of screen scraping and they are blocked anyway if you use them in an img tag)

Pagoda

Answer 4

The URL is a random string as far as I can tell.

It can still be easily retrieved. It is the only img inside the anchor named poster.

So, if you are reading the source, simply search for <a name="poster" and it will be the text following the first src=" from there.

However, you will need to keep the screen scraping code updated because that will probably change.

---

You should also be aware that the images are copyrighted, so be careful to only use the image under a good "fair use" rationale.

Wish

Answer 5

If a thumb is enough, you can use the Facebook Graph API: http://graph.facebook.com/?ids=http://www.imdb.com/title/tt0438097/

Gets you a thumbnail: http://profile.ak.fbcdn.net/hprofile-ak-ash2/50289_117058658320339_650214_s.jpg

Senzer

How did you get that thumbnail url? Has the graph changed since you posted this? Because there's no image URL there now. – Phosphaturia

thats pretty impressive. – Eddy

this was nice, but now you need an access token. further down below however is a simple example how to use themoviedb api (which I have good experiences with) – Senzer

Answer 6

I know that it is way too late, but in my project I used this:-

1. Use omdbapi, Lets take example of Inception, use www.omdbapi.com/?t=inception it will return a json object.
2. In that json object get the "Poster" object, it contains the poster for the image.

Vann

Answer 7

You can use Trakt API, you have to make a search request with the imdb ID, and the Json result given by Trakt API contains links for two images of that movie (poster and fan art) http://trakt.tv/api-docs/search-movies

Hereby

Answer 8

$Movies = Get-ChildItem -path "Z:\MOVIES\COMEDY" | Where-Object {$_.Extension -eq ".avi" -or $_.Extension -eq ".mp4" -or $_.Extension -eq ".mkv" -or $_.Extension -eq<br>  <br>".flv" -or $_.Extension -eq ".xvid" -or $_.Extension -eq ".divx"} | Select-Object Name, FullName | Sort Name <br>
#Grab all the extension types and filter the ones I ONLY want <br>
<br>
$COMEDY = ForEach($Movie in $Movies) <br>
{<br>
        $Title = $($Movie.Name)<br>
        #Remove the file extension<br>
        $Title = $Title.split('.')[0] <br>       
<br>
        #Changing the case to all lower <br>       
        $Title = $Title.ToLower()<br>
<br>
        #Replace a space w/ %20 for the search structure<br>
        $searchTitle = $Title.Replace(' ','%20')       <br>
<br>
        #Fetching search results<br>
        $moviesearch = Invoke-WebRequest "http://www.imdb.com/search/title?title=$searchTitle&title_type=feature"<br>
         <br>
        #Moving html elements into variable<br>
        $titleclassarray = $moviesearch.AllElements | where Class -eq 'title' | select -First 1<br>
<br>
        #Checking if result contains movies<br>
        try<br><br>
        {
            $titleclass = $titleclassarray[0]<br>
        }<br>
        catch<br>
        {<br>
            Write-Warning "No movie found matching that title http://www.imdb.com/search/title?title=$searchTitle&title_type=feature"<br>
        }      <br>
                   <br>
        #Parcing HTML for movie link<br>
        $regex = "<\s*a\s*[^>]*?href\s*=\s*[`"']*([^`"'>]+)[^>]*?>"<br>
        $linksFound = [Regex]::Matches($titleclass.innerHTML, $regex, "IgnoreCase")<br>
         <br><br>

        #Fetching the first result from <br>
        $titlelink = New-Object System.Collections.ArrayList<br>
        foreach($link in $linksFound)<br>
        {<br>
            $trimmedlink = $link.Groups[1].Value.Trim()<br>
            if ($trimmedlink.Contains('/title/'))<br>
            {<br>
                [void] $titlelink.Add($trimmedlink)<br>
            }<br>
        }<br>
        #Fetching movie page<br>
        $movieURL = "http://www.imdb.com$($titlelink[0])"<br>
        <br>
        #Grabbing the URL for the Movie Poster<br>
        $MoviePoster = ((Invoke-WebRequest –Uri $movieURL).Images | Where-Object {$_.title -like "$Title Poster"} | Where src -like "http:*").src  <br> 
<br>
        $MyVariable = "<a href=" + '"' + $($Movie.FullName) + '"' + " " + "title='$Title'" + ">"<br>
        $ImgLocation = "<img src=" + '"' + "$MoviePoster" + '"' + "width=" + '"' + "225" + '"' + "height=" + '"' + "275" + '"' + "border=" + '"' + "0" + '"' + "alt=" +<br> '"' + $Title + '"' + "></a>" + "&nbsp;" + "&nbsp;" + "&nbsp;"+ "&nbsp;" + "&nbsp;" + "&nbsp;"+ "&nbsp;" + "&nbsp;" + "&nbsp;"<br>
        <br>
        Write-Output $MyVariable, $ImgLocation<br>
       <br>
    }$COMEDY | Out-File z:\db\COMEDY.htm  <br>
<br>
    $after = Get-Content z:\db\COMEDY.htm <br>
<br>
    #adding a back button to the Index <br>
    $before = Get-Content z:\db\before.txt<br>
<br>
    #adding the back button prior to the poster images content<br>
    Set-Content z:\db\COMEDY.htm –value $before, $after<br>

Jodeejodhpur

Answer 9

You can use imdb-cli tool to download movie's poster, e.g.

omdbtool -t "Ice Age: The Meltdown" | wget `sed -n '/^poster/{n;p;}'`

Reveille

Answer 10

After playing around with @Hawk's BASE64 discovery above, I found that everything after the BASE64 code is display info. If you remove everything between the last @ and .jpg it will load the image in the highest res it has.

https://m.media-amazon.com/images/M/MV5BMjAwODg3OTAxMl5BMl5BanBnXkFtZTcwMjg2NjYyMw@@._V1_UX182_CR0,0,182,268_AL_.jpg

becomes

https://m.media-amazon.com/images/M/MV5BMjAwODg3OTAxMl5BMl5BanBnXkFtZTcwMjg2NjYyMw@@.jpg

Filmy

so what does MV5BMjAwODg3OTAxMl5BMl5BanBnXkFtZTcwMjg2NjYyMw mean? Should I encode tconst of a movie to base64 what I should I do to get the URL? – Spic

Answer 11

There is one API service provider which will provide you poster image URL and many other details based on the movie name you have provided in their query string.

Over here is the link to the above service provider's website.

You can sign up and use the API service within your code.

Emend

Answer 12

Those poster images don't appear to have any correlation to the title page, so you'll have to retrieve the title page first, and then retrieve the img element for the page. The good news is that the img tag is wrapped in an a tag with name="poster". You didn't say what kind of tools you are using, but this basically a screen scraping operation.

Osmic

The images don't have any correlation to any page simply because one image could be assigned to multiple pages (movies, actors, etc.) and vice verse. – Terrel

Answer 13

Be aware tough, that the terms of service explicitly forbid screenscraping. You can download the IMDB database as a set of text files, but as I understand it, the IMDB movie ID is nowhere to be found in these text files.

Lorusso

Answer 14

I've done something similar using phantomjs and wget. This bit of phantomjs accepts a search query and returns the first result's movie poster url. You could easily change it to your needs.

var system = require('system');

if (system.args.length === 1) {
  console.log('Usage: moviePoster.js <movie name>');
  phantom.exit();
}

var formattedTitle = encodeURIComponent(system.args[1]).replace(/%20/g, "+");
var page = require('webpage').create();
page.open('http://m.imdb.com/find?q=' + formattedTitle, function() {
  var url = page.evaluate(function() {
    return 'http://www.imdb.com' + $(".title").first().find('a').attr('href');
  });
  page.close();
  page = require('webpage').create();
  page.open(url, function() {
    var url = page.evaluate(function() {
      return 'http://www.imdb.com' + $("#img_primary").find('a').attr('href');
    });
    page.close();
    page = require('webpage').create();
    page.open(url, function() {
      var url = page.evaluate(function() {
        return $(".photo").first().find('img').attr('src');
      });
      console.log(url);
      page.close();
      phantom.exit();
    });
  });
});

I download the image using wget for many movies in a directory using this bash script. The mp4 files have names that the IMDB likes, and that's why the first search result is nearly guaranteed to be correct. Names like "Love Exposure (2008).mp4".

for file in *.mp4; do
  title="${file%.mp4}"
  if [ ! -f "${title}.jpg" ] 
    then
      wget `phantomjs moviePoster.js "$title"` -O "${title}.jpg"
  fi
done

Then minidlna uses the movie poster when it builds the thumbnail database, because it has the same name as the video file.

Percipient

Answer 15

Here is my program to generate human readable html summary page for movie companies found on imdb page. Change the initial url to your liking and it generates a html file where you can see title, summary, score and thumbnail.

npm install -g phantomjs

Here is the script, save it to imdb.js

var system = require('system');

var page = require('webpage').create();
page.open('http://www.imdb.com/company/co0026841/?ref_=fn_al_co_1', function() {
  console.log('Fetching movies list');
  var movies = page.evaluate(function() {
    var list = $('ol li');
    var json = []
    $.each(list, function(index, listItem) {
      var link = $(listItem).find('a');
      json.push({link: 'http://www.imdb.com' + link.attr('href')});
    });
    return json;
  });
  page.close();

  console.log('Found ' + movies.length + ' movies');

  fetchMovies(movies, 0);
});

function fetchMovies(movies, index) {
  if (index == movies.length) {
    console.log('Done');

    console.log('Generating HTML');
    genHtml(movies);

    phantom.exit();
    return;
  }
  var movie = movies[index];

  console.log('Requesting data for '+ movie.link);

  var page = require('webpage').create();
  page.open(movie.link, function() {
    console.log('Fetching data');
    var data = page.evaluate(function() {
      var title = $('.title_wrapper h1').text().trim();
      var summary = $('.summary_text').text().trim();
      var rating = $('.ratingValue strong').attr('title');
      var thumb = $('.poster img').attr('src');

      if (title == undefined || thumb == undefined) {
        return null;
      }
      return { title: title, summary: summary, rating: rating, thumb: thumb };
    });

    if (data != null) {
      movie.title = data.title;
      movie.summary = data.summary;
      movie.rating = data.rating;
      movie.thumb = data.thumb;
      console.log(movie.title)
      console.log('Request complete');
    } else {
      movies.slice(index, 1);
      index -= 1;
      console.log('No data found');
    }
    page.close();
    fetchMovies(movies, index + 1);
  });
}

function genHtml(movies) {
  var fs = require('fs');

  var path = 'movies.html';
  var content = Array();

  movies.forEach(function(movie) {
    var section = '';

    section += '<div>';
    section += '<h3>'+movie.title+'</h3>';
    section += '<p>'+movie.summary+'</p>';
    section += '<p>'+movie.rating+'</p>';
    section += '<img src="'+movie.thumb+'">';
    section += '</div>';

    content.push(section);
  });

  var html = '<html>'+content.join('\n')+'</html>';

  fs.write(path, html, 'w');
}

And run it like so

phantomjs imdb.js

Bridgeport

Answer 16

$Title = $($Movie.Name)

$searchTitle = $Title.Replace(' ','%20')  

$moviesearch = Invoke-WebRequest "http://www.imdb.com/search/title?title=$searchTitle&title_type=feature"

$titleclassarray = $moviesearch.AllElements | where Class -eq 'loadlate' | select -First 1

$MoviePoster = $titleclassarray.loadlate

Jodeejodhpur

Answer 17

Now a days, all modern browser have "Inspect" section:

100% Correct for Google Chrome only:

1. Take your cursor on image.
2. Right click on it, select "Inspect Element".
3. In the window appear, under Elements tab you will find the highlighted text as
4. Just click on it.
5. In the Resource tab, right click on image.
6. Select "Copy image URL" option.

Try to paste it any where as URL in any browser, you will only get the image.

Gudgeon