You are brilliant Chris, very clear and detailed presentation!!I use openCV
on a decaptcha .net project, but I did not train it with a set of reference
char as u did, nor do I use KNN. I am using instead the inbuilt OPENCV
Tesseract class along with its set of reference Tessdata files where the
letter vectors are predefined. The code in vb is much more condensed and
quite effective - I have around 50% success rate on 8 letters captcha after
character isolation, rotation and formatting for which I developed code
myself.The vb code is a such.''Dim myOcr as Tesseract = New
Tesseract("","eng",OcrEngineMode.TesseractOnly,"0123456789")myOcr.Recognize(New
Image(Of Bgr, Byte)(bitmap))Dim ocredChar as string = myOct.GetText"Have
you been able to compare the merits of the 2 different approaches?All the
best.
+18F4550videos Thanks for your quick answer and I will look into your license plate demo video asap. I will also let you know the performance & runtime delta between tesseract and KNN on my project when I m done putting them side by side.
Yes, I tried Tesseract with the License Plate Recognition program, demo here: https://www.youtube.com/watch?v=8AIFEN5oODM and found very poor results. The KNN program used here is pretty good if you have good filtering before reading characters, and it would not be especially difficult to make it better or tailor it to a specific use. Also, Tesseract is very difficult to integrate into anything other than Emgu, for example C++ or Python. Tesseract seems like a great project but honestly based on my personal experience I do not recommend it.
Most likely the top of the "i" is being removed due to being too small and not attached to anything, the same as the "." char is removed in the 3.14159. This is a limitation of the current program, I made this with the intention of reading license plates, which generally only have numbers and capital letters. If you need an application to read lower case letters as well it would probably be best to refer to another source as I don't plan on expanding the functionality of this program to include reading lower case letters in the immediate future.
Thanks for the feedback. SVM is most likely the best way to go if matching up an image under test to a database of potentially similar images, I may do a video on this eventually but it could be a while as I have some other videos planned for the immediate future. Stay tuned!
Awesome, but too bad it doesn't work on VS2012 cause of this line :
std::vector intValidChars = { '0', '1', '2', '3', '4', '5', '6', '7',
'8', '9'; }
I see on some forum that you can use this syntaxe only in VS2013.
But nice tutorial ! thank you !
+Jumaan Rimzy It just can't work on a anterior version of VS 2013 ^^Cause the syntax of this line is not implemented in VS 2012 or VS 2010.Just install VS2015 community and it should work. (I didn't try)
+Benjamin Laschkar @ about the correct syntax. I get an error it doesn't work on VS 2010 . Can you give any advice to overcome this problem. Thank you.!
This is an intermediate programming tutorial to create a simple photo blurring tool coded in C++ and utilizing the OpenCV library. Code in the video: ...
Hey phantom awesome it's good love your videos I don't always have time but
your good guy and awesome seeing you have fun with games. You deserve more
but sadly eh well I hope this year more fans and go luck
+Danny tripp You are too kind, man. Here's to 2015 being a year of opportunity for all of us.
Smile recognition using scikit-learn and OpenCV
In this video, I'm using a machine learning algorithm (using the awesome scikit-learn library) trained on 400 faces to detect a smile in the image captured by my ...
Hi friends
I watched all your wideos but l didnt finalized my projects
first step l created GenerateData.cs and l controled my classifications.xml
and ıts lıke that
-
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA==
and l control my images.xml file its llike that
-
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA............AAAAAAAAAA
I thınk thıs reasons my second steps TrainAndTest.cs dont work
when l worked programe ıts select all number from ımage but "number read
from image = " is emty dıdnt wrıte any number to my textbox
please help me why l have like thıs problem
?
+18F4550videos Thank you for your reply First of all l m sure that l m usıng emgucv-windows-universal-cuda 2.4.10.1940 the first GenerateData.cs programme works correctly but ıt forms the classifications.xml and images.xml files incorrectly as you see above...l have a few questions 1 Do l need cuda Nvida graphics card and driver in order to work programme correctly ?2 is there any requirement of library or .xml referance except for Emgu Cv ?Could you please send me OPEN CV KNN Character Recognition CS zip file (complete programme) or images.xml file ? my e-mail is [email protected] or [email protected] ....if l can complete my project l can continue my education life .... its very important for me thank you in advance for your help
Are you sure you were using Emgu 2.4.10? That's what this video was made using. I suggest verifying the install and examples in the following video are working for you:https://www.youtube.com/watch?v=AuCwsHt5a-IIf the concern remains I suggest verifying all the steps have been performed as shown in the video, and if the concern still remains try again on a different computer.I'm going to do an updated version of this with Emgu 3.0 soon, probably within a month or less.
+18F4550videos Thanku so much..!! but i would prefer C# version too if possible! i work in the field of industrial robotics and i think the application of machine learning there would be totally badass!
Thanks for the feedback. I'm currently working on license plate recognition. I have a raw version working in VB but nowhere near ready for presentation yet. I plan on doing C++ and Python versions as well. I'll get it out ASAP but it will be a few weeks at least, stay tuned!
In this tutorial I show how to use the Tesseract - Optical Character Recognition (OCR) in conjunction with the OpenCV library to detect text on a license plate ...
Hi, I've been following your tutorials, and when testing this one, it says
"python: can't open file extract.py'. Maybe python is not compiled? How to
do it?
+Jonecir Souza I found the python file. However, when displaying the output image, I've got the following error:Unable to load font (-helvetica-medium-r-normal--12-iso8859-1) [Resource temporarily unavailable]
The error ValueError: too many values to unpack shows that you use OpenCV
3.0, and findeContours has different syntax.
to fix it change it to:
img_ret, contours, hierarchy = cv2.findContours(edges.copy(),
cv2.RETR_TREE, cv2.CHAIN_APPROX_NONE)
Machine Learning For Remote Sensing : Orfeo ToolBox Meets OpenCV
Orfeo ToolBox is an open-source library developed by CNES in the frame of the Orfeo program since 2006, which aimed at preparing institutional and scientific ...
Help Me!!!
Have a error in this line:
Core.rectangle(mRgba,facesArray[i].tl(),facesArray[i].br(),FACE_RECT_COLOR,3);
the method rectangle(mat point point scalar) is undefined for the type core
+Ostruque10 Yes, since OpenCV 3.0.0, This method is nomoreUse imgproc instead maybe help. I am not sure
NVIDIA Jetson OpenCV Tutorials - Episode 1
Learn to work with mat, OpenCV's primary container. You'll learn memory allocation for a basic image matrix, then test a CUDA image copy with sample ...