foreword
openmv or opencv? I believe that many students who study machine vision have heard of these two words. I myself learned about it when I was preparing for a school competition during the summer vacation. It was originally to use machine vision to identify a certain color or match a certain object, and then communicate with the microcontroller to achieve the purpose of controlling the actions of related modules. So I found some information about computer vision on the Internet, and I will share it with you in this summary, hoping to help your choice.
openMV
openMV is a camera module based on stm32, which can be programmed in Python language, and it comes with many basic image processing algorithms and routines, which is very suitable for machine vision beginners. To put it simply , openMV is a camera processing module with both hardware and software, which is suitable for embedded machine vision and used in conjunction with a single-chip microcomputer. The disadvantage is that the camera pixels are too low, which is not suitable for some complex algorithms such as deep learning, but it is enough for some small projects.
The openMV I used was purchased from Xingtong Technology. There are detailed documentation manuals and teaching videos on the official website. If you read most of the chapters carefully and call the corresponding routines in the IDE to practice and observe, it will be completely finished in two months. Enough to get started, of course, the premise is to have a certain understanding of the Python language (personally feel that Python is really the most suitable programming language for machine vision, and it is easier to learn than other programming languages).
Finally, the official website of Xingtong Technology is attached https://singtown.com/openmv/
openCV
OpenCV, the full name of Open Source Computer Vision Library, is the open source computer vision library , which is the packaged code written by the computer gods and provides corresponding functions. When we use it, we don’t need to write image processing algorithms ourselves, as long as we can Just call the function in its library. It has C++, Python and Java interfaces, and supports Windows, Linux, Mac OS, iOS and Android multiple operating systems. To put it simply , openCV is just an open source vision library, which does not contain any hardware. At the same time, you can also port openCV to embedded development platforms. I don’t know much about it, so you can search for it yourself.
What my own teacher teaches in class is to use Visual Studio C++ as a development tool and call the function library of openCV to realize it. These functions can indeed implement computer vision algorithms and image processing functions very efficiently. openCV has great advantages in solving complex image processing problems. You can do complex projects such as face recognition, motion tracking, human-computer interaction, etc., and you can write your own algorithms to develop more application directions. Due to my limited ability, I only stay at the stage of calling functions, so many of its powerful functions have not been explored, so I can only introduce so much.
Finally, attach the openCV official website https://opencv.org/
end
Finally, to sum up, if you are a beginner and want to participate in competitions or do some small projects, use openMV. If you want to have deeper research in the field of machine vision, learn openCV. My own ability is limited, and I am just a novice who has just started. The content I shared is also where I was confused at the beginning. Now I have accumulated some little experience, and I hope it will be helpful to beginners. Finally, I hope that everyone can persevere on the road of their own learning, not afraid of challenges, and gain something.