NXP S32V23 Design studio and Visual toolbox (MATLAB) Training.

cancel
Showing results for 
Search instead for 
Did you mean: 

NXP S32V23 Design studio and Visual toolbox (MATLAB) Training.

464 Views
Contributor III

Dear All,

We have NXP S32V23 boards and we would like to get started. 

Although the training videos are helpful, we would like to know if there is a workshop where our engineers could have hands-on. This will helps us to get started with the project quickly.

We had sent a request to "Phoenix Micro", but haven't received a reply yet.

Please share the details of the partner companies or agencies, who can provide us training on S32V234.

The following topics are of interest:

1. Usage of Visual Toolbox

2. Usage of Eclipse

3. OpenCV library usage: Build the sample program using OpenCV libraries.

4. How to extends the provide VSD examples, (e.g. ORB, Lanekeeping, etc)

Best regards,

KK

Labels (1)
Tags (3)
10 Replies

7 Views
NXP Employee
NXP Employee

Hi krishnakumar.mayannavar@hs-kempten.de,

In regards with your requests:

 1. Usage of Visual Toolbox

Machine Vision Algorithm development using Vision Toolbox | NXP 

Building Embedded Vision Applications with MATLAB and NXP Vision Toolbox for S32V Video - MATLAB & S... (start with this one for a step-by-step guide)

2. Usage of Eclipse

Please refer to:S32DS for Vision - list of Tutorials 

3. OpenCV library usage: Build the sample program using OpenCV libraries.

Please refer to VisionSDK_UMat.pdf document inside the VSDK documentation package. Here is a snapshot:

vsdk::UMat vsdkumat(HEIGHT, WIDTH, VSDK_CV_8UC3);
cv::UMat cvumat(HEIGHT, WIDTH, VSDK_CV_8UC3);

/* Use of openCV function with combined buffers                 */
/* Explicit retype must be present due to                       */
/* InputArray/OutputArray abstract class in the functions       */
/* No realloc is done here, the blur is done inside of vsdkumat */
cv::blur(cvumat, (cv::UMat)vsdkumat, cv::Size(3, 3));

/* Reading file/buffer via opencv with implicit realloc to OAL  */
vsdk::UMat in = cv::Mat(256, 256, VSDK_CV_8UC1, in).getUMat(cv::ACCESS_READ);
vsdk::UMat in = cv::imread("in.png", 0).getUMat(cv::ACCESS_READ);‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

4. How to extends the provide VSD examples, (e.g. ORB, Lanekeeping, etc)

Just to clarify... you wish to know how to extend the examples delivered with Vision SDK SW package by programming in C++ or you wish to know how to extend/create examples with Matlab by creating new classes in C++ based on Vision SDK ?

Best regards,
Daniel

7 Views
Contributor I

Hi Daniel, 

From the video - Machine Vision Algorithm development using Vision Toolbox I saw a DNN example using SqueezeNet. But I could not find it from the Vision Toolbox installation. How can I get the example, or NXP will put it into a future release? I know SqueezeNet has already been included in the VSDK.

Hongjun

0 Kudos

7 Views
NXP Employee
NXP Employee

Hi Hongjun, 

In the video we showcase what can be done with the NXP Vision Toolbox for S32V from MATLAB as general approaches. The example is going to be available in the next update of the Vision Toolbox what will include support for the latest Vision SDK 1.3.0 too. 

Stay tuned - it will be available in relative short term.

Best regards,

Daniel

0 Kudos

7 Views
Contributor II

Hey All 

Thanks for asking the question... NXP doesn't offer training for this product(only those in the Know). 

I have S32v devices here and done nothing with it just because I couldn't find proper resources.(have it over 9mnths)

I don't have MATLAB. 

Kind regards

Naasik

7 Views
NXP Employee
NXP Employee

Hi Naasik,

Sorry to hear that. Indeed there are not so many other training for S32V product. We did it with MATLAB since that is our team focus. In case you are familiar with MATLAB and you want to give it a try, you can get an evaluation license from MathWorks.

Anyhow, i'm going to forward your complain to the appropiate channels for awarness and who knows ... perhaps that will trigger an change.

Best regards,

Daniel

0 Kudos

7 Views
Contributor II

Hi Daniel

Thanks. I was wondering if I can have a discussion on my use case on a private mail?

Kind regards

Naasik

0 Kudos

7 Views
Contributor II

Hi Daniel

I don't like to use MATLAB as it is very expensive and a trial Licence is no good, if I find a com use case it can't use it due to Licence policy.

For example 1020 and 1050 I had to request the training documentation from my Supplier to get the Training to use the Secure Bootload and correct configuration.

Kind regards

0 Kudos

7 Views
Contributor III

Dear Daniel,

 

Thank you for the reply, it is helpful.

Q4. Is more towards on using the examples delivered with the Vision SDK in terms of its source code, test code, sample data used and API documentation

The video tutorial: NXP\S32DS_Vision_v2.0\S32DS\help\resources\video is helpful, but not sufficient for our requirements. Where can we find related more tutorial videos? 

It would be also helpful if you provide contacts who can provide one or two day(s) hands-on workshop (onsite or off-site/webinar).  This will help our team to a quick start with the project. Let me know your thoughts. 

Regards,

KK

 

0 Kudos

7 Views
NXP Employee
NXP Employee

Histefans‌,

Can you have a look over krishnakumar.mayannavar@hs-kempten.de reply.

Thank you!

Daniel

0 Kudos

7 Views
Contributor III

Hello Stefan Schwarz, Team,

We are waiting for your reply,

We are expected to get the 16 prototypes devices having S32V234 by the end of Aug 2019. At the Phase-1 we are targetting for 300 units of devices. 

Do you or your partner provide an S32V234 workshop? Containing minimum S32V234 vision SDK, Kernel-Graph-Process framework, JPEG decoder, Camera interface, OpenCV/ S32 UMAT usage with few examples and memory allocation/sharing across processing units.

Best regards,

KK

0 Kudos