Macroglossa Visual Search

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Macroglossa
Macroglossa Visual Search Engine Logo, 2012.gif
Web address macroglossa.com
Slogan search is visual
Type of site
Visual Search Engine
Registration optional
Available in English
Created by MVE
Launched 2010
Alexa rank
negative increase 2,828,096 (April 2014)[1]
Current status beta 0.1

Macroglossa is a visual search engine based on the comparison of images,[2][3] coming from an Italian Group. The development of the project began in 2009. In April 2010 is released the first public alpha.[4] Users can upload photos or images that they aren't sure what they are to determine what the images contain. Macroglossa compares images to return search results based on specific search categories. The engine does not use technologies and solutions such as OCR, tags, vocabulary trees. The comparison is directly based on the contents of the image which the user wants to know more.

Interesting features are the categorization of the elements, the ability to search specific portions of the image or start a search from a video file,[5] but the main function is to simulate a digital eye on trying to find similarities of an unknown subject. This feature makes the engine unique.

This technology has several advantages. First, it allows users to pull results from collections of visual content[6] without using tags for search. Second, the visuals can be crowd sourced. In fact by being a search engine, rather than simply a tool, Macroglossa should be able to crowdsourced and scale its recognition vocabulary faster than anyone else and a technology like this would increase the cognitive and spatial skills in humanoid robotics.[7] In addition Macroglosssa can also be used as a Reverse Image Search to find orphan works and possible violations of copyright of images.

Macroglossa supports all popular image extensions such jpeg, png, bmp, gif and video formats such avi, mov, mp4, m4v, 3gp, wmv, mpeg.

Macroglossa enters beta stage in September 2011[8] and at the same time open to the public the opportunity to use the developed interfaces ( Api for web and mobile applications ) in order to expand the use of the engine in the B2B and B2C fields. Macroglossa becomes a SaaS.

API are distributed on three levels : free, basic, and premium. The free API has limited use, but basic and premium do not. The premium API also offers custom services allowing customers to extend and mold the features offered by computer vision.[9]

References[edit]

  1. ^ "Macroglossa.com Site Info". Alexa Internet. Retrieved 2014-04-01. 
  2. ^ Nicola Mattina. "Macroglossa: usare le immagini per effettuare ricerche sul web", Wired.it, Retrieved December 29, 2010.
  3. ^ GreatStartups.com . "Macroglossa.com-What’s In The Picture ", greatstartups.com, Retrieved October 13, 2010.
  4. ^ Liva Judic. "Macroglossa's Visual Search Engine fails to meet basic expectations ", SEW - searchenginewatch, Retrieved April 26, 2010.
  5. ^ Mve. "- Macroglossa PR", - Retrieved 2011.
  6. ^ Make Use OF . "- MacroGlossa: Find Similar Images & Identify Objects In Image ", - Makeuseof.com. 2010.
  7. ^ J. Sturm, A. Visser. "An appearance-based visual compass for mobile robots ", Appearance-based, mobile robot localization, active vision, machine learning. 2000.
  8. ^ Mve. "- macroglossa.com", - Releases and Features. 2011.
  9. ^ J. R. Martínez-de Dios, C. Serna y A. Ollero. "Computer vision and robotics techniques in fish farms ", Robotica. Vo. 21. No. 3. Editor Cambridge University Press. June 2003.

Notes[edit]

  • Wired.it, Retrieved December 29, 2010 : Macroglossa is an Italian project born from a passion for research and innovation by the MVE group of independent developers. The startup has developed a visual search engine based on the comparison of the subjects in the images. The owners of the project define it as "a sort of digital eye can capture, compare and draw conclusions." The purpose of this service is to provide a new type of research within the network. The search engine allows you to upload a picture on the platform and look for similar images on the web. The engine is not based on text tags and does not use OCR to extract strings from images to locate the target. Everything focuses on the key points of the image uploaded by the user. The aim is to give as much information as possible on the results obtained. Each image has a direct result of the source.

External links[edit]