OpenCV installation with CUDA on UBUNTU

hi everyone welcome back today we are gonna install opencv compatible with these packages these libraries which are including the flex in here in the cmake flex before talking about this list maybe we need to do the first steps and we after that we can switch back to the here again first of all you need to download opencv and opencv con trip folders to your directory by these two commands just git colon since these are taking little time according to your internet speed I already downloaded these two folders okay after that you need to specify which opencv version you need to install first of all just go to the opencv contract folder and specify your uh open Siri version since I want to install the latest version which is for act or and I need to get check out to that version okay now it is done for the opencv country folder similarly I need to go to the also opencv folder and doing the same step in here which is gonna be also a git checkout for act all and after that it is done um after that you need to also create a new folder in open series since we need to a build folder we can make our cmake and build inside this folder and after that we can finish our installation at the end okay just enter to the build folder and all is done currently this is the first step you need to do after that the next step is cmx Step which is the most important one because if you are skipping any package any Library here at the end of cmake you are gonna get an error or you are gonna get a get a note which is not found in your uh in your machine these packages this is very important first of all also why we need this list we we need all this for example uh of course we need Cuda because this is also open CB installation specific to compatible with Cuda because I want to use Kuda accelerated functions with the opencv so that's why I need Cuda similarly I need for example and we call it because I want to decode or encode my camera settings or my videos more speedly wire on GPU that's why also I need this package this ambiguity API Also let's say I need FFM pack also because in some cases also I want to um I want to video capture which is familiar which is um famous class in opencv so I want to use it with the flag ffmpeg that's why I need this Library also okay what else also I need with QT enabled because otherwise with IM show I will not be able to use the functionalities the features with my window that's why I need this also and all this in any way I I'm gonna need with open series so that's why I want to install opencv with this Flex enabled okay but if you did not install ffmpeg you are gonna get uh something wrong or other libraries if you have some mistaken if you have some mistakes or if you did something wrong you are gonna get some problems at the end okay just copy this one by the way I'm gonna put this to the descriptions most probably and do these steps in the build folder just enter less little weight I believe it's gonna finish faster I hope so so it's just looking for the packages uh cmx step is just controlling step maybe we can say look at for example Cuda is already detected also you can maybe watch my previous videos about this scooter if you want to install with Cuda also if you don't have Cuda you cannot use this Flex you are gonna maybe get an error because since you don't have GPU you cannot also install Cuda that's why if you um want to install without Cuda you need to check my other videos in my YouTube channel okay let's little check CME is done so configuring done generating is done let's check little bit uh it's already able to find good and also saying the version In Here Also is able to find Cuda also including Enrique and code apis also others kublas others also is able to find where is ffmpic also let's check it also is able to find ffmpeg and with other sub libraries inside is also able to find okay uh seems everything fine we can switch to the next step which is the next step before uh doing 10 maybe we can check our core number since I have 16 cores I can make oh with J 16 so according to your card number you can change here so it can make it faster let's start these steps and I'll since this is little take a long time maybe after that I can continue again [Music] since it's done we can continue the next step is gonna be the installation steps and we are gonna do it with the sudo of course so don't make install is gonna install the opencv to the user local and I think we can start it now it's gonna finish faster I believe okay after it finished we can test the opencv version and we can see it's installed correctly and everything is fine and we can finish the opencv installation I think it's gonna finish faster let's see yes it's done and we can open a new terminal and we can run the command of package config mode version opencv4 and yes so the command of this one is gonna give us the result which is we installed opencv correctly and everything is fine also uh yeah we don't need to make an example maybe for this installation because we are already sure this is installed correctly and yes for this video I wanted to show you how to install opencv compatible with Cuda this is the main content of this video and also maybe uh yeah helpful for you see you in the next videos

Channel: Computer Vision Lab

Views: 1,949

Keywords: opencv, cuda, opencv python, opencv python tutorial, opencv tutorial, opencv installation with cuda, opencv python installation, opencv c++ installation, how to install opencv on ubuntu, how to install opencv with cuda, opencv nvcuvid, opencv cudnn, opencv nvidia, install opencv, opencv 4.8.0, opencv installation in 2023, install latest opencv version, opencv latest release, how to install opencv in 2023, 2023 opencv, opencv install ubuntu, opencv installation with c++, python

Id: 8zXHSfFyXZs

Length: 7min 32sec (452 seconds)

Published: Mon Sep 25 2023