MATLAB VIDEO AND IMAGE PROCESSING BLOCKSET 3 User's Guide download pdf (Page 34)

1 Gettin g Started

Developing Your Models

Use the following g eneral process guidelines to develop real-time video

processing models to run on embedded targets. By o ptimizing the model at

each step, you improve its final p erformance.

1 Create the initial model and optimize the implementation algorithm. Use

floating-point data types so that the model runs faster in simulation mode.

If you are working with a floating-point processor, go to step 3.

2 If you are working w ith a fixed-point processor, gradually change the m odel

data types to fixed point, and run the m odel after e very modification.

During this process, you can use data type conversion b locks to isolate the

floating point sections of the model f rom the fixed-point sections. You should

see a performance improvement if you run the m odel in accelerator mode.

3 Remove unnecessary sink blocks, including scopes, and blocks that log

data to files.

4 Compile the model for deployment on the embedded target.

Code Generation

VideoandImageProcessingBlockset,Real-Time Workshop, and Real-Time

Workshop Embedded Coder enable you to generate code that you can use

to implement your model fo r a practical application. For instance, you can

create an e xecutable from your Simulink m odel to run on a targ et chip. For

more information, see “Understanding Code Generation” in Getting Started

with the Signal P rocessing Blockset.

Windows Dynamic Library Dependencies

To run executables generated for Generic Real-Time (GRT), Embedded

Real-Time (ERT), and S-Function targets, you need

vip_rt.dll if both these

conditions exist:

• The Real-Time Workshop target is a Windows platfo rm.

• You are using the default Real-Time Workshop optimization parameters.

1-26

1 2 ... 29 30 31 32 33 34 35 36 37 38 39 ... 251 252

No comments

MATLAB VIDEO AND IMAGE PROCESSING BLOCKSET 3 User's Guide Page 34