Вы здесь
OpenCV Performance on MAEMO
Опубликовано 28.06.10 пользователем Sergey Balandin
Introduction
OpenCV is an open source effective image processing library. This library expends image and video processing capabilities of Maemo platform and can be used by various applications that process embedded cameras images and videos as well as any other images/videos.. The target of this project is analyzing the possibility of using OpenCV library with MAEMO devices and creating user application that use OpenCV.
Motivation and summary
Augmentation of functionality and of mobile devices allows them to solve tasks before such as image/video processing and pattern recognition. High photo/video cameras resolution and increasing computing power mobile devices graphic acceleration provide the possibility to out some objects (e.g. faces,) in the picture, to do automatic photo processing.
The one of most effective and famous open source image processing libraries is OpenCV. It can perform many of transformations and calculations over images. This library is based on OpenGL and it allows you to improve efficiently of processing by using hardware capabilities of Maemo device.
The main goal of project is OpenCV to Maemo, improving documentation and creating examples as well as testing the functionality of OpenCV library. The prospective project is creating a system of tablet management by fingers. The project includes the realization of a program that will let us manage devices by finger movement before camera. It’s a good illustration of capabilities OpenCV in real-time mode.
Project goals
Today there is no MAEMO port of OpenCV.The tasks of this project are:
-
porting OpenCV to MAEMO
- developing example applications that uses OpenCV
- measuring resource comsumption
Availability of gathered Debian package, documentation and samples of using OpenCV with MAEMO considerably broaden performance capabilities of work with photo/video information and set up premise for user mobile application development.
Project milestones
-
Cross-compilation to ARM
-
Porting to MAEMO4, MAEMO5
-
Test cases
-
Performance testing
-
Performance tuning
-
DEB packaging
-
Project release
Team
- Mironov Ilia
- Diana Ilina
Status:
Graduate
Final deadline:
пятница, апреля 30, 2010 (Весь день)
Users:
