OpenCV maktabad loogu talagalay aqoonsiga sheyga sawirrada iyo kamaradaha

OpenCV waa maktabad bilaash ah oo aragtida mashiinka iskutallaabta ah (noocyada jira ee GNU / Linux, Mac OS X, Windows iyo Android) oo asal ahaan ahaan jiray Waxaa soo saaray Intel loona adeegsaday codsiyo aan tiro lahayn, laga bilaabo nidaamyada amniga ee dhaqdhaqaaqa dhaqdhaqaaqa, si loo farsameeyo codsiyada xakamaynta halka loo baahan yahay aqoonsiga sheyga. Tani waa sababta oo ah daabicitaankeeda waxaa lagu bixiyaa shatiga BSD, kaas oo u oggolaanaya in si xor ah loogu isticmaalo ujeeddooyin ganacsi iyo cilmi baaris iyadoo shuruudaha lagu muujiyey.

Fur CV wuxuu ka kooban yahay in kabadan 500 oo hawlood oo kahadlaya aagag ballaaran oo ah geeddi-socodka aragtida, sida aqoonsiga sheyga (aqoonsiga wajiga), hagaajinta kaamerada, aragtida siteriyaha, aragtida robotka, kala saarida ficilada fiidiyowga, badalida sawirada, soo saarida moodooyinka 3D, abuurista meel 3D ah oo kaamirada isteeriga ah laga dhex abuurayo sawirro tayo sare leh iyadoo la isku darayo sawirrada tayo hooseeya.

Sidoo kale waxay siisaa awood aad ku raadiso sawirrada walxaha la midka ah qaybaha walxaha lagu soo bandhigo iyadoo la adeegsanayo hababka barashada mashiinka, abaabulka astaamaha, aqoonsashada astaamaha guud ee sawirada kala duwan, si toos ah looga takhaluso cilladaha sida indhaha cas.

OpenCV waxay bixisaa in ka badan 2500 algorithms, labadaba kuwo qadiimi ah oo ka tarjumaya guulihii ugu dambeeyay ee laga gaaray dhanka aragtida kumbuyuutarka iyo nidaamyada barashada mashiinka. Koodhka maktabadda waxaa lagu qoray C ++ waxaana lagu qaybiyaa liisanka BSD.

Ku saabsan nooca cusub ee OpenCV 4.2

Hadda maktabaddu waxay ku jirtaa qaabkeedii OpenCV 4.2, taas oo qaybta DNN (Shabakad qoto dheer oo neural ah) oo leh hirgelinta algorithms-ka barashada mashiinka ee ku saleysan shabakadaha neerfaha, wuxuu kudaray gadaal gadaal si loo isticmaalo CUDA iyo taageerada tijaabada ah ee nGraph OpenVINO API ayaa la hirgaliyay.

Isticmaalka tilmaamaha SIMD ka sokow, waxaan ku hagaajineynaa waxqabadka koodhka ee loogu talagalay soo saaridda qalabka (StereoBM / StereoSGBM), cabbirka, maaskarada, isku shaandheynta, xisaabinta qaybaha midabka ka maqan, iyo howlo kale oo badan.

Qeybta G-API (opencv_gapi), kaas oo u shaqeeya sidii matoor wax soo saar Waxtarka muuqaalka iyadoo la adeegsanayo algorithms ku saleysan jaantus, wuxuu taageeraa algorithms isku dhafan oo isku dhafan oo loogu talagalay aragtida kumbuyuutarka iyo barashada mashiinka qoto dheer. Waxay bixisaa taageero Intel Inference Engine. Taageero dheeri ah oo loogu talagalay ka shaqeynta socodka fiidiyowga qaabka fulinta.

Nuglaanshaha (CVE-2019-5063, CVE-2019-5064) taas oo horseedi karta fulinta koodhka weerarka iyadoo laga baaraandegayo xogta aan la hubin ee qaababka XML, YAML iyo JSON sidoo kale waa la hagaajiyay. Haddii astaamo leh null code laga helo inta lagu jiro baaritaanka JSON, qiimaha oo dhan waxaa loo guuriyaa keydka, laakiin iyada oo aan si sax ah loo xaqiijin xuduudaha aagga xusuusta loo qoondeeyay.

Isbeddelada kale lagu soo bandhigay qaybtan cusub:

• Ku darista hirgelinta badan ee waxqabadka pyrDown.
• Waxaa lagu daray awooda laga soosaarayo durdurrada fiidiyowga weelka warbaahinta (demuxing) iyadoo la adeegsanayo Fendate-ka fiidiyowga ku saleysan ee FFmpeg.
• Lagu daray algorithm dib-u-dhiska xulashada soo noqnoqoshada ee sawirrada FSR ee burburay (Dib-u-habeyn Xul Ah)
• Lagu daray habka RIC ee isugeynta meelaha maran ee caadiga ah.
• Lagu daro habka caadiga ah ee leexashada LOGOS.

Sidee loo rakibaa OpenCV 4.2?

Kuwa danaynaya inay awoodaan inay rakibaan maktabaddan, ka heli kartaa nooca cusub sidoo kale la tasho macluumaadka la xiriira isticmaalka oo xitaa ka hel casharro barta rasmiga ah ee rasmiga ah.

Xiriiriyaha waa kan.

Qodobkaan Waxaan kuu soo bandhigi doonnaa tillaabooyinka aan ku hirgelin karno maktabadda on Raspberry pi.

Si aad ugu rakibto OpenCV Raspberry Pwaa inaan haystaa nidaamkaaga, oo ah Raspbian.

Laga soo bilaabo vWaxaan fureynaa terminal gudahana ayaan ku qori doonnaa amarrada soo socda si loo rakibo ku-tiirsanaanta, qalabka horumarinta, baakadaha sawirrada ee maktabadaha kale ee dheeriga ah:

sudo apt-get install build-essential cmake pkg-config
sudo apt-get install libjpeg-dev libtiff5-dev libjasper-dev libpng-dev libavcodec-dev libavformat-dev libswscale-dev libv4l-dev libxvidcore-dev libx264-dev libfontconfig1-dev libcairo2-dev libgdk-pixbuf2.0-dev libpango1.0-dev libgtk2.0-dev libgtk-3-dev libatlas-base-dev gfortran libhdf5-dev libhdf5-serial-dev libhdf5-103 libqtgui4 libqtwebkit4 libqt4-test python3-pyqt5

Ugu dambeyntii aan ku rakibno faylalka madax faylka 3 si aan u soo ururin karno OpenCV:

sudo apt-get install python3-dev

Hadda aan abuurno jawi Python ah amarrada soo socda, tan si loo helo goob go'doon ah:

wget https://bootstrap.pypa.io/get-pip.py
sudo python get-pip.py
sudo python3 get-pip.py
sudo rm -rf ~/.cache/pip

Waxaan rakibeynaa joornaalka iyo virualenvwrapper:

sudo pip install virtualenv virtualenvwrapper
nano ~/.bashrc

# virtualenv and virtualenvwrapper
export WORKON_HOME=$HOME/.virtualenvs
export VIRTUALENVWRAPPER_PYTHON=/usr/bin/python3

source /usr/local/bin/virtualenvwrapper.sh
source ~/.bashrc
mkvirtualenv cv -p python3
pip install "picamera[array]"

tan hadda la sameeyay waxaan ku soo uruurineynaa furanka:

cd ~
wget -O opencv.zip https://github.com/opencv/opencv/archive/4.2.0.zip
wget -O opencv_contrib.zip https://github.com/opencv/opencv_contrib/archive/4.2.0.zip
unzip opencv.zip
unzip opencv_contrib.zip
mv opencv-4.2.0 opencv
mv opencv_contrib-4.2.0 opencv_contrib

Hadda waxaan kordhin doonnaa isku-beddelka nidaamkeenna maxaa yeelay haddii aan uga tagno sida ay tahay markii hore nidaamku wuu sudhan yahay:

sudo nano /etc/dphys-swapfile

Waxaanan wax ka badalneynaa doorsoomaha CONF_SWAPSIZE:

CONF_SWAPSIZE=1024

Waxaan ku keydinaa kuna xirnaa ctrl + o iyo ctrl + x. Kadib waxaan qornaa:

sudo /etc/init.d/dphys-swapfile stop
sudo /etc/init.d/dphys-swapfile start

Hadda waxaan u socon doonnaa soo uruurinta:

workon cv
pip install numpy
cd ~/opencv
mkdir build
cd build
cmake -D CMAKE_BUILD_TYPE=RELEASE \
-D CMAKE_INSTALL_PREFIX=/usr/local \
-D OPENCV_EXTRA_MODULES_PATH=~/opencv_contrib/modules \
-D ENABLE_NEON=ON \
-D ENABLE_VFPV3=ON \
-D BUILD_TESTS=OFF \
-D INSTALL_PYTHON_EXAMPLES=OFF \
-D OPENCV_ENABLE_NONFREE=ON \
-D CMAKE_SHARED_LINKER_FLAGS=-latomic \
-D BUILD_EXAMPLES=OFF ..
make -j4
sudo make install
sudo ldconfig
cd /usr/local/lib/python3.7/site-packages/cv2/python-3.7
sudo mv cv2.cpython-37m-arm-linux-gnueabihf.so cv2.so
cd ~/.virtualenvs/cv/lib/python3.7/site-packages/
ln -s /usr/local/lib/python3.7/site-packages/cv2/python-3.7/cv2.so cv2.so

Oo diyaar.