Skip to content

accelerate ship detection with GPU/CUDA SAR Image Processing

License

Notifications You must be signed in to change notification settings

sty16/ship-detection

Repository files navigation

ship detection using CUDA

GitHub stars GitHub license

1. Compile link method

g++ -c -I/usr/local/MATLAB/R2019b/extern/include/ -I. mat_read.cpp main.cpp
g++ main.o mat_read.o -L/usr/local/MATLAB/R2019b/bin/glnxa64/ -lmat -lmx -leng -lmex -Wl,-rpath /usr/local/MATLAB/R2019b/bin/glnxa64 -o mat

注意:

将-I后的路径更换为MATLAB相应的**/extern/include/路径, 将-L后的路径更换为MATLAB **/bin/glnxa64路径

2. LMM_CFAR folder

This folder contain the CUDA implemetation for ship detection algotithms in synthetic aperture radars images as published inthe following papers.

Yi Cui, Jian Yang, and Yoshio Yamaguchi, “CFAR ship detection in SAR images based on lognormal mixture models,” in Proc. 3rd IEEE Int. Asia-Pacific Conf. on Synthetic Aperture Radar, pp. 1–3, IEEE, Seoul, South Korea (2011).

2.1 Prerequisites

MATLAB r2019b ------>   installed in the /usr/local/MATLAB
opencv4       ------>   installed in the /usr/local/include
CUDA          ------>   installed in the /usr/local/cuda

2.2 make

cd ./LMM_CFAR
make
./LMM_CFAR.out

3. MatToBin Folder

This folder change .mat file to .bin file. The order of data in the binary file: (int)channels (size_t) m (size_t) n (double) data

./matTobin --data ../data/radarsat2-tj.mat -v I -c 1

LICENSE

MIT.