Download #CVriousity 0.2 APK for Android

Name: #cvriousity.apk
ID: com.ramirooliva.cvriousity
Version: 0.2
Size: 0 Mb

#CVriousity Screen Preview

#CVriousity Details

#CVriousity - An OpenCV experiment for Android devices This app has been developed for educational purposes. It's aim is to evaluate OpenCV's ability to detect art imagery using best methods available for current version on Android port. The methods used are aligned to a research in which, from +30 combinations of detectors, descriptors and matchers, a few ones were singled out as best choices (FAST/BRIEF, ORB, GFTT/FREAK plus FAST/FREAK and MSER/FREAK). This app allows you to choose from these selected combinations of methods so you can check their performance and efficacy from your own device and in realtime. Usage: The app allows you to detect objects or images previously analyzed in a scene viewed through device's camera in realtime. You can choose one of the following methods (tap on the menu button on Android device to change active method): - FAST/BRIEF - fastest method / less accuracy. Won't work with rotated images, so you must have the image in portrait mode. - FAST/FREAK - a variation of the previous one, rotation invariant. - ORB - good tradeoff between speed and accuracy. - MSER/FREAK - good accuracy but slower. - GFTT/FREAK- best accuracy but the slowest method of the five. Tap on the screen to focus camera. You can also double tap the screen to freeze the image and let the app try to detect images using each of the available methods. The app comes with 10 art sample images which are automatically added to the catalog (database) the first time the app is run. You may also take your own pictures of small objets using device's builtin camera. I recommend you to put a blank, white sheet, below the object before taking the photo. The app will automatically crop and resize image before adding it to the catalog. To test the app you can download reference images from: https://www.dropbox.com/sh/0pqe5vzr557hrsu/pmpmA1Wkee A reference sheet for the images can be downloaded from https://www.dropbox.com/sh/0pqe5vzr557hrsu/pmpmA1Wkee Individual images can be viewed online at Web Gallery of Art (http://www.wga.hu) Art: Image 1 - "Allegory", AACHEN, Hans von http://www.wga.hu/art/a/aachen/allegory.jpg Image 2 - "Bacchus, Ceres and Cupid", AACHEN, Hans von, http://www.wga.hu/art/a/aachen/bacchus.jpg Image 3 - "Joking Couple", AACHEN, Hans von, http://www.wga.hu/art/a/aachen/j_couple.jpg Image 4 - "The Archangel Michael", ABADIA, Juan de la, http://www.wga.hu/art/a/abadia/michael.jpg Image 5 - "Albarello", ABAQUESNE, Masséot, http://www.wga.hu/art/a/abaquesn/albarell.jpg Image 6 - "Ceramic Floor", ABAQUESNE, Masséot, http://www.wga.hu/art/a/abaquesn/floor1.jpg Image 7 - "Ceramic Floor", ABAQUESNE, Masséot, http://www.wga.hu/art/a/abaquesn/floor2.jpg Image 8 - "The Flood", ABAQUESNE, Masséot, http://www.wga.hu/art/a/abaquesn/theflood.jpg Image 9 - "Chimney breast", ABBATE, Niccolò dell', http://www.wga.hu/art/a/abbate/chimney1.jpg Image 10 - "Chimney breast", ABBATE, Niccolò dell' (2) http://www.wga.hu/art/a/abbate/chimney2.jpg Point the camera to them and try different detection methods. Notes: The app is 100% free source software. See LICENSE for licensing details. The official repository of this software is: https://github.com/RamiroOliva/CVriousity.git Please, feel free to contact me (ramiro_oliva@hotmail.com or @RamiroOliva) if you have any questions.

#CVriousity | 9 Reviewers | 2013/6/1 12:00:00 SA | Rating: 4.3