/image-hunter

Smart image scraper, spider and image depot builder.

Primary LanguagePython

Image-Hunter: The New Generation Image Retrieval Engine

Image-Hunter is a personal AI project that incorporates search engines, web spiders, autonomous agents, machine learning, self-learning, and natural language processing. The system has evolved into a search engine with new features that enable users to search the web based on images. The HTML interface displays the images and some of the web content, allowing users to visit the source of the images to find more images or specific articles. The project is evolving very quickly, and soon development will be transferred to the search engine project.

Hard-Core Development and Web Interface

The script automatically opens a browser showing the image results in different views. Search queries can be executed with the terminal by simply entering a query or phrase, and the system retrieves the images and some information related to the search query. The interface is currently in development and will use JavaScript to provide an advanced image viewing interface. The developer is conducting research on JavaScript libraries that have image viewers integrated into them. The software will be packaged, and installation documentation will be prepared for quick installation of the scraper-spider.

Image-Miner Features and Functionality

Image-Hunter is the new name for Image-Miner, and it works with Google, Yahoo, and Bing. It learns from website and user input and checks links, images alt tag, image filename, and web page path. The system has a simple scoring system called QS (Quality Score), extracts and reads HTML content, has a simple HTML interface, a database to keep track of information, builds a depot of all images, and keeps a history. The config.env file allows users to adjust the parameters of the scraper.

Web Server and Supported Image Formats

A web server version is being implemented, and the software supports various image formats, including PSD, APNG, AVIF, BMP, GIF, ICO, CUR, TIF, TIFF, JPG, JPEG, JFIF, PJEG, PJP, PNG, SVG, WEBP, WEBM, OGG, WAV, TIFF, ICO, JPG, GIF, PNG, and BMP. Database integration is completed with SQLite, and during testing in the past four weeks, the software gathered 85,000 images and icons.

Future Developments

The developer is continuously working to add more intelligence to the system. In 2022, the developer's wish list includes search library optimization, automatic search results, archiving development, DB development, vision development, enhancing AI, smarter search algorithms, and better roaming and crawling.

Next Phase: Judging the Quality of Images

The next phase of Image-Hunter will focus on judging the quality of an image. The system already does an impressive job with no visual information, and when the vision algorithm is ready, the software will do an amazing job.

Conclusion

Image-Hunter is an innovative open-source image search engine that utilizes AI and other related fields to provide the best image search results. The system is constantly evolving, and the developer is working hard to add more features and functionality to the software. With its simple interface, support for various image formats, and ability to learn from website and user input, Image-Hunter is a valuable tool for anyone searching for images on the web. The software is available on GitHub, and users can adjust the parameters to fit their needs for optimal performance.