🤖 Overview

PictureFinder is a web crawler operated by Picsearch AB (Stockholm, Sweden), as documented on their official website picsearch.com and the crawler policy page at “http://www.picsearch.com/bot.html”. Its primary purpose is to discover and index publicly accessible images from web pages for the Picsearch image search engine, which provides visual search results to partner websites and powers features like reverse image lookup. The bot was first observed in 2006 and continues to crawl the web for new and updated visual content, respecting all robots.txt directives.

🌐 Technical Behavior

PictureFinder crawls the web using a headless browser emulation to render JavaScript-heavy pages and extract image URLs from dynamic content. It sends an average of one request per second per domain, with occasional burst requests during initial discovery, and obeys any Crawl-Delay directive found in robots.txt. The bot’s IP ranges are owned by Picsearch’s hosting provider and are listed in public DNS records; typical ranges include 185.53.xxx.xxx (IPv4) and 2a01:4f8:xxx:xxxx::/64 (IPv6). It supports HTTP/1.1 and HTTP/2 protocols and sends conditional GET requests with If-Modified-Since headers to reduce bandwidth usage. The crawler parses XML sitemaps to efficiently discover image URLs and checks for the “nofollow” attribute on links to avoid unwanted paths. It also extracts image metadata such as file size, dimensions, and alt text from the surrounding HTML.

📋 robots.txt Compliance

PictureFinder fully honors robots.txt directives, as stated on the Picsearch crawler policy page. The bot will not crawl any path disallowed by the site owner, and it respects the noindex meta tag for images. Evidence from multiple webmaster forums confirms that blocking PictureFinder via robots.txt (e.g., “User-agent: PicsearchBot Disallow: /images”) effectively stops its crawling. The company also provides a dedicated email ([email protected]) for site owners to request immediate removal from their index.

🔍 Detection Indicators

The primary User-Agent string is “PicsearchBot/1.0 (; http://www.picsearch.com/bot.html)” but the bot also uses the string “PictureFinder/1.0” in some versions. Behavioral fingerprints include a low request rate (typically 1–3 seconds between requests per domain) and the absence of typical browser headers like Accept-Language or Sec-Fetch-* . The bot adds a custom header “From: [email protected]” and often includes a trailing slash in its User-Agent. Server logs show requests for image file extensions such as jpg, png, gif, webp, and svg, often with a Referer header pointing to the parent page. The bot’s IPs can be looked up via reverse DNS entries ending in “picsearch.com”.

📊 Data Usage

Collected image metadata (URL, alt text, surrounding text, file dimensions, and checksums) is used to populate the Picsearch image index. This data also feeds into training image recognition algorithms and improving search relevance for visual queries. According to picsearch.com, the company does not sell user data or share it with third parties, and they remove images upon site owner request via the “Remove My Image” tool on their website.

⚙️ Rate Limiting Policy

Although legitimate, PictureFinder is rate-limited because its crawling, even at moderate rates, can cause performance degradation on low-resource or poorly optimized servers. A threshold-based block at 5 requests per second per IP or 100 requests per minute is recommended to balance accessibility with server stability, as documented in common web application firewall (WAF) rule sets.

ⓘ Data Notice: The information presented above has been compiled from publicly available internet sources. Boteraser aggregates this data solely for informational purposes and does not independently classify, evaluate, or endorse any findings about the bots listed. The accuracy and completeness of this information is the sole responsibility of the original publishers. Boteraser and its operators accept no liability for any decisions made based on this data.