Creating a Photo Mosaic from League of Legends Skin Images Using Python Web Scraping

The article introduces the concept of a "photo mosaic" where a large image is composed of many smaller pictures, and describes the author's inspiration from a League of Legends (LOL) skin mosaic they saw online.

Origin : The author discovered a mosaic made from nearly a thousand LOL skin screenshots and decided to recreate it by first collecting all skin images.

Crawling approach : Because the LOL skin gallery uses a paging system that does not change the URL, the author inspected the network requests (F12) to find the URL pattern for each skin image. The pattern is https://game.gtimg.cn/images/daoju/app/lol/medium/2-<heroID><skinID>-9.jpg , where heroID identifies the champion and skinID is a three‑digit number (001‑015).

By extracting hero IDs from the JavaScript file champion.js and iterating over possible skin numbers, the script builds a complete list of image URLs.

<code>import requests</code>
<code>import re</code>
<code>import os</code>
<code># title:获取LOL英雄皮肤图像</code>
<code># author:简书 Wayne_Dream</code>
<code># date:2018-7-5</code>
<code>def getHero_data():</code>
<code>    try:</code>
<code>        headers = {'user-agent':'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/67.0.3396.99 Safari/537.36'}</code>
<code>        url = 'http://lol.qq.com/biz/hero/champion.js'</code>
<code>        r = requests.get(url, headers=headers)</code>
<code>        r.raise_for_status()</code>
<code>        r.encoding = r.apparent_encoding</code>
<code>        text = r.text</code>
<code>        hero_id = re.findall(r'"id":"(.*?)","key"', text)</code>
<code>        hero_num = re.findall(r'"key":"(.*?)"', text)</code>
<code>        return hero_id, hero_num</code>
<code>    except:</code>
<code>        return 'Failed to get hero IDs!'</code>
<code>def getUrl(hero_num):</code>
<code>    part1 = 'https://game.gtimg.cn/images/daoju/app/lol/medium/2-'</code>
<code>    part3 = '-9.jpg'</code>
<code>    skin_num = []</code>
<code>    Url_list = []</code>
<code>    for i in range(1, 21):</code>
<code>        i = str(i).zfill(3)</code>
<code>        skin_num.append(i)</code>
<code>    for hn in hero_num:
<code>        for sn in skin_num:
<code>            part2 = hn + sn
<code>            url = part1 + part2 + part3
<code>            Url_list.append(url)
<code>    print('图片URL获取成功')
<code>    return Url_list
<code>def PicName(hero_id, path):
<code>    pic_name_list = []
<code>    for id in hero_id:
<code>        for i in range(1, 21):
<code>            pic_name = path + id + str(i) + '.jpg'
<code>            pic_name_list.append(pic_name)
<code>    return pic_name_list
<code>def DownloadPic(pic_name_list, Url_list):
<code>    count = 0
<code>    n = len(Url_list)
<code>    try:
<code>        for i in range(n):
<code>            headers = {'user-agent':'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/67.0.3396.99 Safari/537.36'}
<code>            res = requests.get(Url_list[i], headers=headers).content
<code>            if len(res) < 100:
<code>                count += 1
<code>                print('\r当前进度：{:.2f}%'.format(100*(count/n)), end='')
<code>            else:
<code>                with open(pic_name_list[i], "wb") as f:
<code>                    f.write(res)
<code>                    count += 1
<code>                    print('\r当前进度：{:.2f}%'.format(100*(count/n)), end='')
<code>    except:
<code>        return 'Failed to download images!'
<code>if __name__ == '__main__':
<code>    print('author:简书 Wayne_Dream:')
<code>    print('https://www.jianshu.com/u/6dd4484b4741')
<code>    input('请输入任意字符开始爬虫:')
<code>    if not os.path.exists('D:\\LOLimg_wayne\\'):
<code>        path = r'D:\\LOLimg_wayne\\'
<code>        os.mkdir(path)
<code>        hero_id, hero_num = getHero_data()
<code>        Url_list = getUrl(hero_num)
<code>        pic_name_list = PicName(hero_id, path)
<code>        print('正在下载图片，请稍等。。。')
<code>        print('在' + path + '下查看...')
<code>        DownloadPic(pic_name_list, Url_list)
<code>        print('图片下载完毕')
<code>    else:
<code>        path = r'D:\\LOLimg_wayne\\'
<code>        hero_id, hero_num = getHero_data()
<code>        Url_list = getUrl(hero_num)
<code>        pic_name_list = PicName(hero_id, path)
<code>        print('正在下载图片，请稍等。。。')
<code>        print('在' + path + '下查看...')
<code>        DownloadPic(pic_name_list, Url_list)
<code>        print('图片下载完毕')

The script creates the destination folder, fetches hero IDs, builds all possible image URLs, generates local filenames, and downloads each image while showing progress.

Mosaic creation : After obtaining the full set of skin pictures, the author uses a free mosaic software (downloadable from https://fmedda.com/en/download) to import the images, select "Create photo mosaic", and generate the final large picture. Screenshots of the software UI and the resulting mosaic are provided.

Finally, the author shares a Baidu Cloud link containing the complete skin image collection for readers who want to skip the crawling step.

Original article link: https://www.jianshu.com/p/c963370cd8df