wanguu / libpabod Goto Github PK
View Code? Open in Web Editor NEWThis project forked from mjmarin/libpabod
libPaBOD: a C++ library for PArt Based Object Detection on images.
License: MIT License
This project forked from mjmarin/libpabod
libPaBOD: a C++ library for PArt Based Object Detection on images.
License: MIT License
libpabod: Library for PArt-Based Object Detection in C++ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Daniel Rodriguez Molina and Manuel J. Marin-Jimenez This software implements the object detection system described in Felzenszwalb et al. [1]. Contents of the package: ======================== - bin - the executable file will be created here - data: - models - contains all the object model files - testimages - a set of images to test the software - doc - folder where the documentation will be generated - include - contains all the software header files - libs - folder where the library will be generated - obj - folder where the object files will be generated during the compilation - src - contains all the software source files - tests - contains some programs to test the library - makefile - used to compile the library, the documentation and the test program Requirements: ============= This software has been tested on Ubuntu 10.10 (Maverick Meerkat) with the following libraries: - libmatio-dev - v1.3.4-2 (required) - libcv-dev - v2.1.0-2 (required) - libhighgui-dev - v2.1.0-2 (required) - doxygen - v1.7.1-1 (optional; used to generate the documentation) - graphviz - v2.26.3-4 (optional; used to generate figures for the documentation) Quick start: ============ A) Classic installation (obsolete --> CMake is highly recommended): 1. Unpack in <libpabod_directory> 2. cd <libpabod_directory> 3.1. Generating the library > make 3.2. Creating the test programs > make alltests 3.3. Generating the documentation > make docum 3.4. Recompiling all > make cleanobj > make Tip: use 'make all' to generate both the library and the test programs. B) Installing with CMake: Please, follow the steps described in the file named INSTALL. C) Testing the library: (assuming test program has been already generated) 1. cd <libpabod_directory> 2. ./bin/detectobj -m <model_path> -i <image_path> [-t <threshold>] [-o <detections_path>] [-d <0/1>] Example: > ./bin/detectobj -m data/models/person_v6.mat -i data/testimages/2008_007537.jpg -t -0.3 -o detections.txt If the program has finished correctly, you will find a text file named 'detections.txt' with the following structure: <number_of_detected_objects> <x1_i> <y1_i> <x2_i> <y2_i> <score_i> where <x1_i> <y1_i> <x2_i> <y2_i> are the pairs of coordinates of the i-th bounding box and <score_i> is its corresponding detection score. Tip: object detection can be performed directly on video frames (e.g. avi file) with the test program named 'detectvid'. Citation: ========= If you use this library for your publications, please cite it as: @misc{libpabod, author = {Rodriguez-Molina, Daniel and Marin-Jimenez, Manuel J.}, title = {{LibPaBOD}: A Library for Part-Based Object Detection in {C++}}, year = {2011}, note = {Software available at \url{http://www.uco.es/~in1majim/}} } Contact the authors: ==================== Daniel Rodriguez Molina (developer) - [email protected] / [email protected] Manuel J. Marin-Jimenez (advisor) - [email protected] References: =========== [1] P. Felzenszwalb, R. Girshick, D. McAllester, D. Ramanan. "Object Detection with Discriminatively Trained Part Based Models." IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 32, No. 9, September 2010. URL: http://people.cs.uchicago.edu/~pff/latent/ Version history: ================ - v0.2.7: memory leaks found by github/yokox solved. Thanks! - v0.2.6: new file 'export.h' for properly exporting functions to dll's. - v0.2.5: updated CMake files and other minor improvements for Windows compatibility. - v0.2.4: CMake and code updated to properly manage libmatio on Windows (thanks to Eric Sommerlade for his invaluable contributions). See new file 'INSTALL.windows'. libmatio 1.5.x is supported. - v0.2.2: detector component is now included in detections matrix. Other minor improvements. - v0.2.1: OpenMP is now optional via CMake. Selection of OpenMP overrides pthread's choice. - v0.2: new class Pabod. It encapsulates the class model and detection process. New demo file 'detectobj2.cpp' uses Pabod class. pthread is now optional via CMake. New 'quickstart.pdf' in directory 'doc'. - v0.1.7: fixed bugs (i.e. memory leaks) reported by <anonymous>. - v0.1.6: fixed documentation option for CMake (thanks to eichnerm). Use 'cmake .. -DINSTALL_DOC=ON" to generate the library documentation. - v0.1.5: CMake is supported (thanks to [email protected]). By default, the library is built in shared mode (non-static). - v0.1.4: added an alternative 'makeDetection' function. Updated makefile. Now 'make all' builds both the library and the test programs. - v0.1.3: added 'detectvid' program. It performs detection on video sequences (e.g. avi). Use 'make test3' to generate it. Use 'make alltests' to generate all test programs. - v0.1.2: small fixes in code. - v0.1.1: added 'detectobj' program. It can save detections to disk. Use 'make test2' to generate it. - v0.1: first release.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.