VisionSDK_ISS_SensorFrameWork Vision SDK User Guide Opencx

User Manual:

Open the PDF directly: View PDF .
Page Count: 8

Download
Open PDF In Browser	View PDF

Vision SDK
(v03.xx)
OpenCX

Copyright © 2014 Texas Instruments Incorporated. All rights reserved.
Information in this document is subject to change without notice. Texas Instruments may have pending
patent applications, trademarks, copyrights, or other intellectual property rights covering matter in this
document. The furnishing of this documents is given for usage with Texas Instruments products only and
does not give you any license to the intellectual property that might be contained within this document.
Texas Instruments makes no implied or expressed warranties in this document and is not responsible for
the products based from this document.
Page 1 of 8

IMPORTANT NOTICE
Texas Instruments and its subsidiaries (TI) reserve the right to make changes to their products or to
discontinue any product or service without notice, and advise customers to obtain the latest version of relevant
information to verify, before placing orders, that information being relied on is current and complete. All
products are sold subject to the terms and conditions of sale supplied at the time of order acknowledgment,
including those pertaining to warranty, patent infringement, and limitation of liability.
TI warrants performance of its products to the specifications applicable at the time of sale in accordance with
TI’s standard warranty. Testing and other quality control techniques are utilized to the extent TI deems
necessary to support this warranty. Specific testing of all parameters of each device is not necessarily
performed, except those mandated by government requirements.
Customers are responsible for their applications using TI components.
In order to minimize risks associated with the customer’s applications, adequate design and operating
safeguards ought to be provided by the customer so as to minimize inherent or procedural hazards.
TI assumes no liability for applications assistance or customer product design. TI does not warrant or represent
that any license, either express or implied, is granted under any patent right, copyright, mask work right, or
other intellectual property right of TI covering or relating to any combination, machine, or process in which such
products or services might be or are used. TI’s publication of information regarding any third party’s products or
services does not constitute TI’s approval, license, warranty or endorsement thereof.
Reproduction of information in TI data books or data sheets is permissible only if reproduction is without
alteration and is accompanied by all associated warranties, conditions, limitations and notices. Representation
or reproduction of this information with alteration voids all warranties provided for an associated TI product or
service, is an unfair and deceptive business practice, and TI is neither responsible nor liable for any such use.
Resale of TI’s products or services with statements different from or beyond the parameters stated by TI for
that product or service voids all express and any implied warranties for the associated TI product or service, is
an unfair and deceptive business practice, and TI is not responsible nor liable for any such use.
Also see: Standard Terms and Conditions of Sale for Semiconductor Products.
www.ti.com/sc/docs/stdterms.htm
Mailing Address:
Texas Instruments
Post Office Box 655303
Dallas, Texas 75265

Copyright © 2014, Texas Instruments Incorporated

Page 2 of 8

TABLE OF CONTENTS
1
1.1
1.2
1.3
1.4

OpenCV: Introduction .................................................................................. 4
Building cmem ................................................................................................. 4
Steps to load cmem kernel module in Linux ......................................................... 4
Running OpenCV test-bench on Vision SDK BIOS ................................................. 4
Cloning and Building TI OpenCV source ............................................................... 5

2
2.1
2.2
2.3

OpenCL: Compiling OpenCL Kernels ............................................................. 6
Linux .............................................................................................................. 6
SYSBIOS ......................................................................................................... 6
Compilation of OpenCL kernels of Vision SDK use-cases ........................................ 6

3

Revision History ........................................................................................... 8

Page 3 of 8

1

OpenCV: Introduction
OpenCV provides open compute library and OpenCL offers open standard for
heterogeneous programming. Both these open components are enabled in Vision
SDK. This document describes steps needed for OpenCL to work in Linux Vision SDK.

1.1

Building cmem
In order to build cmem, follow the instructions in the below link
http://processors.wiki.ti.com/index.php/CMEM_Overview
The git link and checkout
‘VisionSDK_Linux_UserGuide’.

tag

of the

cmem repo are mentioned

in

the

In the above wiki, refer to the ‘Build Environment Setup’ and ‘Building Test Binaries’
sections to set up the build environment to build cmem user and kernel modules.
While configuring the makefile generation use the below command instead of the one
mentioned in the wiki so that shared libraries are generated instead of static
libraries.
./configure --host=arm-linux-gnueabihf --prefix=$PWD
The build process remains same otherwise.

1.2

Steps to load cmem kernel module in Linux
For OpenCL to work in Linux, cmem kernel module is required. Cmem provides a
contiguous range of memory to OpenCL device. So, we need to load this kernel
module after the system has booted up.
Once Linux Vision SDK has booted up, navigate to the below folder
$> cd /opt/vision_sdk
Then execute the script ‘load_cmem_ko.sh’ to load the cmem kernel module.
$> ./load_cmem_ko.sh
Now, run usual steps described in VisionSDK_Linux_UserGuide.pdf, as below
$> ./camnodes.sh
$> ./vision_sdk_load.sh
$> ./apps.out
Select the demo to run

1.3

Running OpenCV test-bench on Vision SDK BIOS
In order to run OpenCV test-bench from the BIOS environment, a Vision SDK
usecase is added.

Page 4 of 8

1.3.1

1.3.2

Enabling OpenCV test-bench:
Follow the below steps to enable OpenCV test-bench:


Enable building the OpenCV test-bench use-case by setting the make
variable
ENABLE_OPENCV_TESTS
to
yes
in
cfg.mk
of
tda2xx_evm_bios_opencx/cfg.mk file



The OpenCV test-bench is built as library and linked to this Vision SDK
usecase

Configuring the test-bench
In order to configure the test-bench do the following:


The test-bench entry function names are similar to its counterpart
executable names. E.g opencv_test_imgproc.exe could be run by
calling test_imgproc() and setting the appropriate args in the
algplugin.



Similarly, for opencv_perf_imgproc.exe, the counterpart function entry
point would be perf_test_imgproc()



Next download the test data from the below link
o



1.4

https://github.com/opencv/opencv_extra

Copy the testdata folder to the SD card before running any opencv test
that reads any file from the testdata folder

Cloning and Building TI OpenCV source
TI OpenCV source repository contains works related to


Enabling cross build on ARM A15 running BIOS or Linux



Offload of certain functions from A15 to C66x DSP



Test bench changes to enable running in BIOS environment

This repository can be cloned with the below command
git clone https://bitbucket.itg.ti.com/scm/opencv/tiopencv.git
In order to build opencv source as part of vision sdk build, follow the below
procedures


Navigate to ‘ti_components’ folder



Then, navigate to ‘algorithms_codecs’



Now, clone the ‘tiopencv’ repo using the above mentioned command



Set the env variable, ‘BUIILD_OPENCV_SRC’
‘tda2xx_evm__opencx’ config



Now, executing make with the target ‘opencx’ will build vision_sdk and
also opencv for the appropriate configuration and the opencv build
contents will be placed in one of the below folders inside ‘tiopencv’
folder
o

build_bios_debug

o

build_bios_release

o

build_linux_debug

o

build_linux_release

to

‘yes’

in

the

In order to clean, use ‘opencx_clean’ target with make.

Page 5 of 8

Note: If BUILD_OPENCV_SRC is set to ‘yes’, always use the target opencx to build
and clean

2

OpenCL: Compiling OpenCL Kernels
In this section we describe the steps for compiling the OpenCL CL kernels with Vision
SDK. The Open CL kernels are compiled with the compiler clocl. Clocl is in available
at \packages\ti\clocl. More details about
clocl and OpenCL in general are available at
http://downloads.ti.com/mctools/esd/docs/opencl/index.html

2.1

Linux
The following steps are to be followed for offline compilation of OpenCL Kernels.
1. Set your system Path to //OpenCL_Installation_Folder/packages/ti/clocl.
2. Set the following variable to
a. TARGET_ROOTDIR = //YOUR_TARGET_FS/
3. Now you can compile your CL files.

2.2

SYSBIOS
In case of sysbios the CL files must be compiled on a host Linux machine.

2.2.1

Basic Compilation of Kernels:
This step is valid when there is no change in the DSP image from the standard Vision
SDK release.
1. Set your system Path to //OpenCL_Installation_Folder/packages/ti/clocl.
2. Set TARGET_ROOTDIR = //vision_sdk_Installation_Folder/examples/tda2xx/src/opencl

3. Now you can compile your CL files.
4. If you are including the CL files as pre-complied header then now you have to
build the vision SDK.
2.2.2

Advanced Steps:
These steps are required if there is any change in the DSP 1 Code or Tools or any
new component is added or there is a memory map change.
1. Do a clean build of Vision SDK on a linux host.
2. Copy dsp.syms and dsp.syms.obj from
//vision_sdk_Installation_Folde/binaries/tda2xx_evm/bios/opencx/vision_sdk/bin/tda2xxevm to

//vision_sdk_Installation_Folder/examples/tda2xx/src/opencl/usr/share/ti/opencl
3. Now follow the Basic Compilation steps described in section 4.2.1.
4.

2.3

Compilation of OpenCL kernels of Vision SDK use-cases

In order the compile the OpenCL kernels of Vision SDK use-cases, the following are
required to be done.

Page 6 of 8

2.3.1

Setting up Environment variables
Set up the following environment variables before building vision sdk
1. The following two steps are needed only if opencl kernel is compiled without
vision sdk
a. Add clocl path to PATH variable. (this is already taken care in vision
sdk)
i. This can be found inside opencl folder inside ti_components for
BIOS and
ii. Inside /usr/share/ti/opencl/bin/x86 for Linux
b. Set TI_OCL_CGT_PATH variable pointing to dsp compiler root path
(this is already taken care in vision sdk)
2. Add dsp compiler bin path to PATH variable
a. i.e. add the compiler binary path to PATH variable
b. e.g export PATH=/bin:$PATH

2.3.2

BIOS:
Follow the below steps:
3. The OpenCL kernels should reside inside the folder named ‘kernel’ of the
respective algplugin
4. A Makefile is required to be present inside this folder and this can be just
copied from one of the existing examples. E.g. from ‘openclframecopy’
algplugin
5. Once the above two steps are done, building vision sdk with target named
‘opencx’ is required which builds these OpenCL kernels along with usual Vision
SDK build
a. i.e. make opencx –s –j
6. Note: if there is no intention of building OpenCV do
BUILD_OPENCV_SRC to yes in ‘tda2xx_evm__opencx’ config

2.3.3

not

set

Linux:
Follow the below steps:
1. The first two steps mentioned in BIOS section, 2.3.1 are to be followed
2. Then, add a build target named ‘opencl_build’ to the MAKEFILE.MK file of the
algplugin
3. Then, in addition to performing make on the algplugin MAKEFILE.MK default
target, perform make on the algplugin MAKEEFILE.MK ‘opencl_build’ target
a. This
is
done
from
the
‘MAKEFILE_adas.MK’
/hlos/examples/ folder

found

in

Page 7 of 8

3

Revision History
Version

Date

Revision History

1.0

18th October 2016

1.1

25 October 2016

1.2

9 June 2017

1.3

5 July 2017

1.4

th

th

th

22

nd

March 2018

Initial Version
Addressed review comments
Added steps to build OpenCV
from source
Minor corrections in path and
footer
Updated OpenCV repo path

««« § »»»

Page 8 of 8



---

Source Exif Data:

File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.5
Linearized                      : No
Page Count                      : 8
Language                        : en-US
Tagged PDF                      : Yes
Title                           : VisionSDK_ISS_SensorFrameWork
Author                          : TI
Creator                         : Microsoft® Word 2010
Create Date                     : 2018:04:06 12:58:48+05:30
Modify Date                     : 2018:04:06 12:58:48+05:30
Producer                        : Microsoft® Word 2010

EXIF Metadata provided by EXIF.tools