🤖 Overview

imagevisu is a web crawler operated by ImageVisu GmbH, a company headquartered in Berlin, Germany, that powers a visual search engine and image recognition platform. The bot was first documented in 2018 and its primary purpose is to systematically collect publicly accessible images from websites to build a proprietary visual index used for reverse image search, brand monitoring, and AI-driven object detection training. The product, also called ImageVisu, offers a commercial API for businesses to identify logos, products, and scenes in user-uploaded images.

🌐 Technical Behavior

The crawler makes HTTP GET requests predominantly for image files (JPEG, PNG, GIF, WebP) and also scrapes the parent HTML pages to extract alt-text and surrounding metadata. It respects standard crawl delays by defaulting to a maximum of 10 requests per second per domain, though this rate can spike during initial deep crawls of large sites. Requests originate from IP addresses registered under the AS20773 Host Europe GmbH range, with a dedicated subnet block 185.104.184.0/24 verified through WHOIS lookup. The bot uses HTTP/1.1 with the Accept header set to `image/webp,image/*,*/*;q=0.8` and includes a From header containing a contact email ([email protected]). It follows redirects up to five hops and caches DNS lookups for one hour to reduce server load.

📋 robots.txt Compliance

According to the official documentation published at https://imagevisu.com/robots-txt-policy, the imagevisu bot fully honors Disallow directives in robots.txt files. The crawler re-checks the robots.txt file every 24 hours and will immediately cease crawling any URL path blocked via a Disallow rule. However, it does not interpret the Allow directive in older robots.txt specifications, relying strictly on RFC 9309 compliance. There are no reported incidents of the bot ignoring explicit crawl disallowances.

🔍 Detection Indicators

The primary User-Agent string is imagevisu (case-sensitive, lowercase), sometimes appearing as ImageVisu/1.0 or Mozilla/5.0 (compatible; ImageVisu/1.0; +https://imagevisu.com/bot). A secondary fingerprint is the presence of the Via header bearing the value 1.1 imagevisu-proxy when requests pass through a caching layer. The bot also sets a custom HTTP header X-ImageVisu-Crawl-ID during testing phases, though this is omitted in production. Log analysis from major CDNs shows a clear pattern of requests concentrated on pages containing <img> tags with non-empty src attributes.

📊 Data Usage

Collected images are processed using convolutional neural networks to extract feature vectors and store them in a high-dimensional search index. This index powers the ImageVisu reverse image search API, which businesses use for copyright enforcement, product identification, and brand protection. Additionally, a subset of the crawled data (approximately 2% of total images) is used to train proprietary object detection models, such as the ImageVisu LogoNet, which can recognize over 10,000 brand logos with 94% accuracy as reported in the company’s 2023 technical whitepaper.

⚙️ Rate Limiting Policy

Due to its continuous crawling of image-rich resources, imagevisu can generate significant bandwidth consumption, making rate-limiting essential. The recommended policy is to throttle requests to a maximum of 5 requests per second per IP after observing the bot’s default burst pattern, and to block any request exceeding 20 requests per second to prevent resource exhaustion without permanently denying legitimate access.

ⓘ 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.