Image Searching - does it work, what are tips to get the best images, how does it work:

Image searching has always been limited by the text-based nature of most search engines. If it's not text, it won't be found. However, various entities are working on a solution in the form of sketch-based search engines. The user simply draws what he or she is looking for, and the engine retrieves instances of similar geometry. The first commercial offering is CADFind Sketch and Search from Applied Search Technology of the United Kingdom. It finds 2D engineering drawings based on a user sketch or existing drawing. According to the company, it offers retrieval accuracy of 95% for the most basic drawing. It's currently limited to searching DXF files. The master program, used to code part properties to create a searchable catalog, costs $850 for a single license. Clients with search-only capability are free (Advanstar Communications, 2004).

Karthik Ramani of Purdue University (http://tools.ecn. purdue.edu/~cise/dess.html) created a system that can find computer-designed industrial parts based on a sketch of what you're looking for. If a search result appears close, you can manipulate a sketch of that part to bring it closer to what you have mind, then search again. The search system can be used as a stand-alone feature for CAD repositories or in conjunction with PLM systems. The search engine works by converting stored CAD models and user-inputted sketches to voxels, the 3D equivalent of a pixel. It then compares voxel patterns for similarities (Advanstar Communications, 2004).

Princeton University professor Thomas Funkhouser and colleagues developed a 3D search engine that lets the user to use his mouse to sketch 2D or 3D object, add a text description, and then search for similar models. There's also an option to upload your own file and search for similar models. VRML V2.0 is the only supported file type, but other formats such as DXF and Unigraphics UG usually work if they contain only geometry.
Image searching is sometimes offered by general search engines, like Google or Yahoo!, but there are also specialized image search engines - services devoted to indexing images or multimedia. In addition, there are meta- search engines, which pass on search requests to more than one search engine and then bring back the results. Sometimes 'Image Search Engine' is also used to refer to collection-based search engines - services that index a single or small number of image collections. Large digital libraries or commercial stock photo collections, like Corbis, typically offer their own search engine-like facilities. All of these types of image search engines are text-based - their indexes are created from words associated with the images. In addition, there have been attempts to create content-based search engines, which 'index' visual characteristics of an image, such as its shape and color. However, these attempts are still largely experimental and are often limited to single image collections. Image search engines are based on existing search engine technology, but they use additional strategies to identify, categorize and rank images. A search engine's indexing of images is done automatically, rather than using human indexers, so it must find ways to guess at the image's content. It might take into account its filename or any accompanying 'ALT' picture tags (these are coded into the HTML page). It might look for clues from the image's context - for example, the words or phrases that are close to the image, or the 'META' tags found at the top of the HTML coding. The nature of the Web site and its provider may also be taken into account.

Analysis of an image's text and context can be used to exclude images as well as include them - for example, an image engine will usually consider an image's context and associated words when it is blocking out adult material. The Google and Yahoo! are the major image searching engines. Yahoo! has acquired its own sophisticated search engine technology and now supplies image search functionality to Lycos, Altavista and AllTheWeb. AOL Search draws its image results from Google, and Ask Jeeves uses PicSearch. Hotbot and MSN Search do not offer true image search features: they simply enable users to limit their search results to 'only pages with images'.

Google claims to search 800 million images and 4 billion Web sites. It's important to note that although large numbers of results are reported, Google will not actually allow its users to view more eight or nine hundred images. The results are presented as thumbnails first which link to a split screen with Google on top and the source page below. Information includes: filename, extension, pixel dimensions, file size and the location URL. This is accompanied with a simple copyright warning. Usefully, Google groups together like pictures from the same site, presenting one to the user and offering a link to 'more'. This can make for a better set of results.

Yahoo! shows a mere one third results compared to Google. Results can be limited by size, file type, color or Web domain, with an adult material filter turned 'on' or 'off'. Once results are retrieved there are no options to further filter the search. Results are quick, images are good and relevant, no dead links or duplicates were observed. Yahoo! does not actually allow its users to view more than about 1100 images. The results are displayed in the form of thumbnails first: a fixed 20 at a time, presented within a 115x115 pixel slide frame. These links to a split screen with Yahoo! on top and the source page below. Information includes: filename, extension, pixel dimensions, file size and the location URL. This is accompanied (in the split-screen view) with a simple copyright warning.

There are also 3 specialized image search engines: Cobion visoo, Ditto and Picsearch. Cobion Visoo has vanished Cobian still offer an image search as part of the German search engine Dino (http://www.dino-online.de/). Cydral is a newcomer, but offers very limited results and much of its Web site is in French. Of these three specialized image search engines, Picsearch scores the best. Picsearch is owned by a Swedish company. In addition to its own dedicated site, it licenses its image search to Ask Jeeves. Images can by limited by color, animation and pixel dimensions. Adult material is automatically excluded. Results are generally very good and relevant. No duplication or dead links were observed. The declared results may be smaller than Google or Yahoo! but the user can actually view many more of them. The display of results is good with 16 thumbnails first and then links to a split window with Picsearch on top and the source below. Information includes pixel dimensions, file size, file type and color info - including number of colors in the image. There is a clear statement about copyright and no sign of advertising.

Image search engines attempt to give access to the wide range of images available on the Internet. Generally speaking, the results of the large automated image search engines are disappointing in quality. The frequent irrelevancy of their results is also explicable, since the automated engines are guessing at their images' visual subject content using indirect textual clues. The ability to filter a search - to include and exclude items - is important in any Web searching, but particularly so when searching for images. Many users will wish to exclude adult imagery from their search results, but it can also be very useful to limit by file type, file size, or color - and the ability to use Boolean logic or phrases will greatly improve the relevancy of the results. Meta- image search engines fail their users. They are inadequate in providing quality pictures.