Embedded Image Processing on the TMS320C6000™ DSP

hi_china59

Embedded Image Processing on the TMS320C6000™ DSP: Examples in Code Composer Studio™ and MATLAB focuses on efficient implementations of advanced image processing algorithms for resource-constrained embedded DSP systems. Featuring the popular Texas Instruments TMS320C6000TM line of Digital Signal Processors, it utilizes a proven methodology of beginning with a high-level algorithmic point-of-view, proceeding to develop prototype code in MATLAB and Visual Studio, and concluding with an efficient implementation suitable for deployment on an embedded DSP system. Aside from image processing, the author also explains the uses and rationale behind a plethora of technologies, most notably several industry-standard and essential TI developer technologies, including the Code Composer Studio™ IDE. Highlights include numerous debugged MATLAB and C/C++ Visual Studio prototype applications and efficient C implementations of real-world algorithms tested on both the C6416 DSK and C6701 EVM development platforms. Packaged with the book is a CD-ROM which includes the source code, tested and debugged using the latest versions of developer tools, as well as sample image data. The CD-ROM contains a number of projects, spanning a diverse range of image processing applications such as, spatial processing techniques, image filtering, edge detection and segmentation schemes, and wavelet-based algorithms, among others. Embedded Image Processing on the TMS320C6000™ DSP: Examples in Code Composer Studio™ and MATLAB is an essential book for professional signal & image processing engineers working with TI DSPs where real-time constraints are present and performance is at a premium. Imaging software developers and DSP users will also find this book applicable, as it covers a variety of image and signal processing building blocks that appear in a diverse set of real-world applications, including medical imaging, satellite imaging, digital photography, and pattern recognition, to name a few. It may also serve as a reference work for advanced image processing, computer vision, and DSP students working in labs that use TI development kits or MATLAB.

Acknowledgements xvii

2. Tools 15
2.1. The TMS320C6000 Line of DSPs 16
2.1.1 VLIW and VelociTI 17
2.1.2 Fixed-Point versus Floating-Point 21
2.1.3 TI DSP Development Tools
(C6701 EVM & C6416 DSK) 25
2.2 TI Software Development Tools 26
2.2.1 EVM support libraries 28
2.2.2 Chip Support Library 28
2.2.3 DSP/BIOS 29
2.2.4 FastRTS 29
2.2.5 DSPLIB and IMGLIB 29
x Contents
2.3 MATLAB 30
2.4. Visual Studio .NET 2003 30
2.4.1 Microsoft Foundation Classes (MFC) 31
2.4.2 GDI+ 33
2.4.3 Intel Integrated Performance Primitives (IPP) 34
3. Spatial Processing Techniques 37
3.1 Spatial Transform Functions and the Image
Histogram 37
3.2 Contrast Stretching 40
3.2.1 MATLAB Implementation 45
3.2.2 TI C67xx Implementation and MATLAB Support Files 47
3.3 Window/Level 59
3.3.1 MATLAB Implementation 63
3.3.2 A Window/Level Demo Application Built
Using Visual Studio .NET 2003 68
3.3.3 Window/Level on the TI C6x EVM 77
3.4 Histogram Equalization 82
3.4.1 Histogram Specification 87
3.4.2 MATLAB Implementation 90
3.4.3 Histogram Specification on the TI C6x EVM 94
4. Image Filtering 103
4.1 Image Enhancement via Spatial Filtering 103
4.1.1 Image Noise 106
4.1.2 2D Convolution, Low-Pass and High-Pass Filters 109
4.1.3 Fast Convolution in the Frequency Domain 111
4.1.4 Implementation Issues 114
4.2 Linear Filtering of Images in MATLAB 116
4.3 Linear Filtering of Images on the TI C62xx/C67xx 118
4.3.1 2D Filtering Using the IMGLIB Library 120
4.3.2 Low-Pass Filtering Using DSPLIB 124
4.3.3 Low-Pass Filtering with DSPLIB and Paging 128
4.3.4 Low-Pass Filtering with DSPLIB and Paging via DMA 132
4.3.5 Full 2D Filtering with DSPLIB and DMA 138
4.4 Linear Filtering of Images on the TI C64x 139
4.4.1 Low-Pass Filtering with a 3x3 Kernel Using IMGLIB 141
4.4.2 A Memory-Optimized 2D Low-Pass Filter 146
4.5 Non-linear Filtering of Images 152
4.5.1 Image Fidelity Criteria and Various Metrics 152
4.5.2 The Median Filter 155
4.5.3 Non-Linear Filtering of Images in MATLAB 159
4.5.4 Visual Studio .NET 2003 Median Filtering Application 166
4.5.4.1 Generating Noise with the Standard C Library 167
4.5.4.2 Profiling Code in Visual Studio .NET 2003 169
4.5.4.3 Various Median Filter C/C++ Implementations 171
4.5.5 Median Filtering on the TI C6416 DSK 179
4.6 Adaptive Filtering 183
4.6.1 The Minimal Mean Square Error Filter 185
4.6.2 Other Adaptive Filters 187
4.6.3 Adaptive Image Filtering in MATLAB 190
4.6.4 An MMSE Adaptive Filter Using the Intel IPP Library 194
4.6.5 MMSE Filtering on the C6416 198
5. Edge Detection and Segmentation 211
5.1 Edge Detection 212
5.1.1 Edge Detection in MATLAB 219
5.1.2 An Interactive Edge Detection Application with
MATLAB, Link for Code Composer Studio, and RTDX 225
5.1.2.1 DSP/BIOS 229
5.1.2.2 C6416 DSK Target 232
5.1.2.3 C6701 EVM Target 236
5.1.2.4 Host MATLAB Application 240
5.1.2.5 Ideas for Further Improvement 245
5.2 Segmentation 246
5.2.1 Thresholding 248
5.2.2 Autonomous Threshold Detection Algorithms 249
5.2.3 Adaptive Thresholding 254
5.2.4 MATLAB Implementation 257
5.2.5 RTDX Interactive Segmentation Application with
Visual Studio and the TI C6416 261
5.2.5.1 C6416 DSK Implementation 262
5.2.5.2 Visual Studio .NET 2003 Host Application 269
6. Wavelets 281
6.1 Mathematical Preliminaries 282
6.1.1 Quadrature Mirror Filters and Implementing the
2D DWT in MATLAB 287
6.1.2 The Wavelet Toolbox 296
6.1.3 Other Wavelet Software Libraries 299
6.1.4 Implementing the 2D DWT on the C6416 DSK
with IMGLIB 299
6.1.4.1 Single-Level 2D DWT 301
6.1.4.2 Multi-Level 2D DWT 305
6.1.4.3 Multi-Level 2D DWT with DMA 309
6.2 Wavelet-Based Edge Detection 313
62A The Undecimated Wavelet Transform 316
6.2.2 Edge Detection with the Undecimated
Wavelet Transform 317
6.2.3 Multiscale Edge Detection on the
C6701 EVM and C6416 DSK 323
6.2.3.1 Standalone Multiscale Edge Detector (C6701 EVM) 323
6.2.3.2 HPI Interactive Multiscale Edge Detector Application
with Visual Studio and the TIC6701 EVM 329
6.2.3.2.1 C6701 EVM Target 331
6.2.3.2.2 Visual Studio .NET 2003 Host Application 334
6.2.3.3 Standalone Multiscale Edge Detector (C6416 DSK) 341
6.3 Wavelet Denoising 347
6.3.1 Wavelet Denoising in MATLAB 352
6.3.2 Wavelet Denoising on the C6x 359
6.3.2.1 D4 DWT and EDWT functions on the C6416 362
6.3.2.2 A C6416 Wavelet Denoising Implementation 374
Appendix A Putting it Together: A Streaming Video
Application 379
A. 1 Creation and Debugging of MEX-files
in Visual Studio .NET 2003 382
A. 1.1 The import_grayscale_image MEX-file 385
A. 1.2 A MEX-file for HPI communication between
MATLAB and the C6x EVM 390
A.2 The C6701 EVM Program 393
A.3 MATLAB GUI 396
A.4. Ideas for Further Improvement 398
Appendix B Code Optimization 401
B.I Intrinsics and Packed Data Processing 404
B. 1.1 Packed Data Processing 405
B. 1.2 Optimization of the Center of Mass Calculation
on the C64x Using Intrinsics 408
B.2 Intrinsics and the Undecimated Wavelet Transform 415
B.3 Convolution and the DWT 418

[ 本帖最后由 hi_china59 于 2009-10-5 12:38 编辑 ]

apteye

xiexie

mllw

下到后面不出来了，估计是没有信元了，这样分包怎么也撑不住啊！

elf_chillymoon

谢谢分享 正好要用

fighter5555

4# elf_chillymoon

zzr1974418

haha-----太奢侈了.到这里XIA
http://www.aybook.cn/book/16038.html

gongping

好人呀！！！！！！！！！！！！

lcj2212916

好书啊！！！谢谢

lcj2212916

不错的书，谢谢楼主了

lcj2212916

好书！！！！！！！！！支持一下