Texas Instruments | Canny Edge Detection Implementation on TMS320C64x/64x Using VLIB | Application notes | Texas Instruments Canny Edge Detection Implementation on TMS320C64x/64x Using VLIB Application notes

Texas Instruments Canny Edge Detection Implementation on TMS320C64x/64x  Using VLIB Application notes
Application Report
SPRAB78 – November 2009
Canny Edge Detection Implementation on
TMS320C64x/64x+ Using VLIB
Y Senthil Kumar ...............................................................................................................................
ABSTRACT
This application note contains instructions for implementing Canny Edge Detection algorithm using VLIB.
The general ideas given here can also be applied for an application which calls a sequence of VLIB
functions in a block-based manner.
1
2
3
4
5
6
Contents
Introduction ..................................................................................................................
Canny Edge Detector Algorithm ..........................................................................................
2.1
Noise Reduction ....................................................................................................
2.2
Gradient Filtering ...................................................................................................
2.3
Non-Maximum Suppression ......................................................................................
2.4
Hysteresis Thresholding ..........................................................................................
Wrapper Function ...........................................................................................................
3.1
Frame Processing ..................................................................................................
3.2
Block Processing ...................................................................................................
Experiments and Results ..................................................................................................
Conclusion ...................................................................................................................
References ...................................................................................................................
2
2
2
3
3
4
5
5
5
6
6
6
List of Figures
1
Flow Diagram of Canny Edge Detector .................................................................................. 2
2
The Sequence of Block Outputs After Each Stage ..................................................................... 5
3
The Sequence of Image Outputs After Each Stage .................................................................... 6
List of Tables
1
Performance Figures on DM6437 EVM .................................................................................. 6
SPRAB78 – November 2009
Submit Documentation Feedback
Canny Edge Detection Implementation on TMS320C64x/64x+ Using VLIB
Copyright © 2009, Texas Instruments Incorporated
1
Introduction
1
www.ti.com
Introduction
Canny edge detector is the optimal and most widely used algorithm for edge detection. Compared to other
edge detection methods like Sobel, etc canny edge detector provides robust edge detection, localization
and linking. It is a multi-stage algorithm and the stages involved are illustrated in Figure 1. Thus, instead
of providing the whole algorithm as a single API, kernels are provided for each stage. This way, the user
can have more flexibility and better buffer management. For more details on the theory and formulation
please refer to the Canny edge detection paper [1].
Optimized kernels and wrappers required for the implementation of canny algorithm are provided to
reduce the efforts of the integrator. This document is to aid the integrator to use the kernels, understand
the algorithm implemented and use the appropriate parameters. Basic image processing and C64x
programming knowledge are assumed. For a formal description of the APIs for these kernels, refer to the
VLIB documentation (Section 16 through Section 19) [3].
Gaussian Smoothing
Gradient Filtering
Non-maximum Suppression
Hysteresis Thresholding
Figure 1. Flow Diagram of Canny Edge Detector
2
Canny Edge Detector Algorithm
The kernels involved in the canny edge detector algorithm are discussed in detail in this section. At each
stage, for computing the output pixel at a particular row, we need input pixels at rows below and above.
Thus, output at the first and the last row are undefined. The same happens in the case of columns too.
Thus, the size of the valid pixels in the output reduces after each step. To incorporate this, the output
width and height and the output buffer’s position changes after each step. This is illustrated in each stage
as API argument adjustments. This API adjustment technique is the same for other applications where a
sequence of VLIB functions are called.
2.1
Noise Reduction
Noise brings about high gradient magnitudes which in turn produces unintended edges. To reduce this
effect, the image is convolved with a 2D Gaussian filter, which brings each pixel in closer harmony with its
neighbors. This is basically a smoothing process. Practically, a discrete Gaussian function requires a
window size of 5*sigma approximately (where sigma is the standard deviation of Gaussian function).
Sigma is the parameter which controls the smoothing, which in turn boils down to window size. Larger
window sizes are generally not recommended because it is computationally very expensive and also
causes over smoothing, removing weak edges. A window size of 5x5 or 7x7 is recommended. The
Gaussian filters can be generated using the matlab command fspecial.
For the implementation of this step on the C64x, optimized 2D convolution functions
IMG_conv_3x3_i8_c8s, IMG_conv_5x5_i8_c8s and IMG_conv_7x7_i8_c8s can be used depending on the
window size (conv_window_size) required. Due to the edge effects of convolution, the output image’s
height and width decreases by (conv_window_size-1). Care is to be taken in using the modified height and
width for the next step according to the conv_window_size used. Please refer to IMGLIB documentation
[2] for more details on the usage of the above convolution functions.
2
Canny Edge Detection Implementation on TMS320C64x/64x+ Using VLIB
Copyright © 2009, Texas Instruments Incorporated
SPRAB78 – November 2009
Submit Documentation Feedback
Canny Edge Detector Algorithm
www.ti.com
The required API argument adjustments are illustrated in Equation 1. The parameter conv_window_size is
the window size of the convolution kernel.
pBufIn = input;
pBufOut = putput + width * (conv _ window _ size - 1) / 2 + (conv _ window _ size - 1) / 2;
pGx = pBufGradX + width * (conv _ window _ size - 1) / 2 + (conv _ window _ size - 1) / 2;
pGy = pBufGradY + width * (conv _ window _ size - 1) / 2 + (conv _ window _ size - 1) / 2;
pMag = pBufMag + width * (conv _ window _ size - 1) / 2 + (conv _ window _ size - 1) / 2;
output _ height = height - (conv _ window _ size - 1);
output _ width = width - (conv _ window _ size - 1);
(1)
API Usage:
IMG_conv_7x7_i8_c8s( pBufIn,
pBufOut,
output_width* output_height,
width,
gaussian_7x7,
8 );
It is to be noted that Gaussian filter is a spatially separable kernel. Thus the 2D convolution can be
performed as 1D row and column convolutions which saves conv_window_size/2 computations for each
output calculation. But for smaller window sizes (<7) this does not give significant improvement in
performance on the C64x.
2.2
Gradient Filtering
This kernel computes the first order horizontal and vertical gradients, and computes the magnitude of the
gradient using L1 norm. This step creates an additional 1-pixel border of invalid data around the input
image. As before, care has to be taken to use the modified height and width for the next step. The
calculation steps used in the algorithm are listed in Equation 2.
Gx = I(x + 1, y) - I(x - 1, y)
Gy = I(x, y + 1) - I(x, y - 1)
Gmag = (| Gx | + | Gy |)
(2)
The required API argument adjustments are illustrated in Equation 3.
pGx+ = ( width + 1);
pGy + = ( width + 1);
pMag + = ( width + 1);
output _ height - +2;
(3)
API Usage:
VLIB_xyGradientsAndMagnitude( pBufOut,
pGx,
pGy,
pMag,
width,
output_height );
2.3
Non-Maximum Suppression
The edge magnitude image may contain wide ridges around the local maxima which are visualized as
thick edges. Non-maximum suppression produces thin edges removing the non-maxima pixels along the
normal direction preserving the connectivity of contours. At each pixel location, two virtual points lying
along the normal direction on either side of the current location are interpolated using the gradient
magnitudes from surrounding neighbors. If the gradient magnitude of the current position is greater than
those at the virtual points, it is declared a possible edge by giving a value of 127. Else it is made 0. The
direction estimation and interpolation steps were combined to eliminate division, and hence, removing the
reciprocal tables required for division. Finally, the output is binary valued containing either 0 or 127. This
step creates another 1-pixel border of invalid data around the image.
SPRAB78 – November 2009
Submit Documentation Feedback
Canny Edge Detection Implementation on TMS320C64x/64x+ Using VLIB
Copyright © 2009, Texas Instruments Incorporated
3
Canny Edge Detector Algorithm
www.ti.com
The required API argument adjustments are illustrated in Equation 4.
pGx+ = ( width + 1);
pGy + = ( width + 1);
pMag + = ( width + 1);
pBufOut + = ( 2 * width + 2);
output _ width - = 4;
output _ height - = 2;
(4)
API usage:
VLIB_nonMaximumSuppressionCanny( pMag,
pGx,
pGy,
pBufOut,
output_width,
width,
output_height );
2.4
Hysteresis Thresholding
This stage of the algorithm is split into a block based (VLIB_doublethresholding) and a non-block based
(VLIB_edgeRelaxation) kernel. This is done to give the flexibility of using a part of this stage in a block
based manner. Double thresholding uses two parameters highthreshold and lowthreshold. If the gradient
magnitude is above high threshold and it is a possible edge, then it is declared a strong edge and a value
of 255 is stored. Its location is stored in a dynamic list where the size of the list grows based on the edge
structure.
The API usage is illustrated below. Buffer listptr holds the positions of strong edges. A buffer used for pGx
or pGy can be re-used for this buffer.
VLIB_doublethresholding( pMag,
pBufOut,
listptr,
numItems,
output_width,
width,
output_height,
loThres,
hiThres
0);
The second kernel traverses each strong edge and updates the list with a weak edge if it is a neighbor.
Thus, it is not block based as it traverses through the image arbitrarily. The performance of this kernel
depends on the number of strong edges and weak edges, as well as their organization. Usually, the
percentage of edges in an image is small, thus this kernel runs for less time.
VLIB_edgeRelaxation( pBufOut,
listptr,
numItems,
width );
4
Canny Edge Detection Implementation on TMS320C64x/64x+ Using VLIB
Copyright © 2009, Texas Instruments Incorporated
SPRAB78 – November 2009
Submit Documentation Feedback
Wrapper Function
www.ti.com
3
Wrapper Function
Two implementations of wrapper function for Canny are provided; namely, Frame based and block based
processing implementations.
3.1
Frame Processing
The first implementation contains a straight forward usage of the kernels. As it is not complicated, it is
easier to understand and get started. This implementation is suitable for small images that can fit in the L1
memory. Code segments shown before as API adjustments were taken from
VLIB_testCannyEdgeDetector.c, which is present in INSTALL_DIR\VLIB_V_2_1\examples\src.
3.2
Block Processing
For larger images, all the buffers will not fit L1 memory. Thus, we have to load blocks of the image into L1
memory from DDR and do the processing of each stage. We provide a BlockBasedCanny.c file to
illustrate this block based processing of Canny Edge Detector Algorithm. The first four kernels are used in
a block based manner. The last kernel VLIB_edgeRelaxation cannot be implemented in a block based
manner. Thus, it is run on the whole image.
The first three kernels reduce the output size by 2 each (for conv_window_size=3). Thus, seven lines of
input are required for the computation of one line of output. From these seven lines of input, we obtain five
lines of Gaussian smoothed output, three lines of Gradient filtered output and one line of non-maximum
suppressed output. This method is required only for the first output line computation. For the computation
of the next output line, we need the next seven lines where the first six lines were used for previous
computation. By trying to use the previous computations, we can reduce the computation for the next
output line and figure out a steady state scheme.
In steady state, we need to allocate only four lines of memory for input, Gaussian smoothed output,
Gradient filtered output and Non-maximum suppressed output buffers. The scheme is illustrated in
Figure 2. In each iteration, four lines of input are used to compute two lines of Gaussian smoothed output
and then the last two lines shaded in red are moved to the first two lines of the input buffer for the next
iteration. Now for the next iteration, only two new input lines are loaded to the last two lines. Similarly in
the Gaussian smoothed output buffer, two lines are reused from previous iteration and two lines are
computed. The same logic is extended for the subsequent stages.
This block based version of the code can be easily plugged into a ping-pong buffering scheme using
EDMA drivers.
Figure 2. The Sequence of Block Outputs After Each Stage
In Figure 2, the sequence of block outputs after each Stage: four rows of Input block, Gaussian Smoothed
block, Gradient Filtered block and Non-maximum suppressed block. In steady state, all these steps
produce four rows of output. Red filled rows are re-used for the next iteration.
SPRAB78 – November 2009
Submit Documentation Feedback
Canny Edge Detection Implementation on TMS320C64x/64x+ Using VLIB
Copyright © 2009, Texas Instruments Incorporated
5
Experiments and Results
4
www.ti.com
Experiments and Results
Canny edge detector algorithm explained as above was run on the DM6437 EVM and the results after
each step is shown in Figure 3. The performance numbers for each kernel and its parameters involved are
tabulated below.
Figure 3. The Sequence of Image Outputs After Each Stage
In Figure 3, the sequence of image outputs after each Stage: Input Image, Gaussian Smoothed Image,
Gradient Filtered Image, Non-maximum suppressed Image, Final Edge Output
Table 1. Performance Figures on DM6437 EVM
Function
Cycles Per Pixel
Parameters
VLIB_gaussianFilter7X7
1.5(3x3), 4.25 (5x5), 7.5(7x7)
conv_window_size
VLIB_xyGradientsAndMagnitude
0.8
VLIB_nonMaximumSuppressionCanny
8.7
VLIB_doublethresholding
2.4
VLIB_edgeRelaxation
(1)
5
0.6-3.6
High and low threshold
(1)
Input dependent
Conclusion
The implementation of canny edge detector using VLIB was discussed. Simulation and benchmarking was
performed on the 64x+ simulator and the functionality was tested on DM6437, DM642 and DM6446.
6
References
1. A Computational Approach to Edge Detection by Canny, J., IEEE Trans. Pattern Analysis and Machine
Intelligence, 8:679-714, 1986.
2. TMS320C64x+ DSP Image/Video Processing Library (v2.0.1)
3. Vision Library (VLIB) Application Programming Interface Reference Guide (SPRUG00). Describes the
VLIB Application Programming Interface (v2.1)
6
Canny Edge Detection Implementation on TMS320C64x/64x+ Using VLIB
Copyright © 2009, Texas Instruments Incorporated
SPRAB78 – November 2009
Submit Documentation Feedback
IMPORTANT NOTICE
Texas Instruments Incorporated and its subsidiaries (TI) reserve the right to make corrections, modifications, enhancements, improvements,
and other changes to its products and services at any time and to discontinue any product or service without notice. Customers should
obtain the latest relevant information before placing orders and should verify that such information is current and complete. All products are
sold subject to TI’s terms and conditions of sale supplied at the time of order acknowledgment.
TI warrants performance of its hardware products to the specifications applicable at the time of sale in accordance with TI’s standard
warranty. Testing and other quality control techniques are used to the extent TI deems necessary to support this warranty. Except where
mandated by government requirements, testing of all parameters of each product is not necessarily performed.
TI assumes no liability for applications assistance or customer product design. Customers are responsible for their products and
applications using TI components. To minimize the risks associated with customer products and applications, customers should provide
adequate design and operating safeguards.
TI does not warrant or represent that any license, either express or implied, is granted under any TI patent right, copyright, mask work right,
or other TI intellectual property right relating to any combination, machine, or process in which TI products or services are used. Information
published by TI regarding third-party products or services does not constitute a license from TI to use such products or services or a
warranty or endorsement thereof. Use of such information may require a license from a third party under the patents or other intellectual
property of the third party, or a license from TI under the patents or other intellectual property of TI.
Reproduction of TI 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. Reproduction of this information with alteration is an unfair and deceptive
business practice. TI is not responsible or liable for such altered documentation. Information of third parties may be subject to additional
restrictions.
Resale of TI 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 and is an unfair and deceptive business practice. TI is not
responsible or liable for any such statements.
TI products are not authorized for use in safety-critical applications (such as life support) where a failure of the TI product would reasonably
be expected to cause severe personal injury or death, unless officers of the parties have executed an agreement specifically governing
such use. Buyers represent that they have all necessary expertise in the safety and regulatory ramifications of their applications, and
acknowledge and agree that they are solely responsible for all legal, regulatory and safety-related requirements concerning their products
and any use of TI products in such safety-critical applications, notwithstanding any applications-related information or support that may be
provided by TI. Further, Buyers must fully indemnify TI and its representatives against any damages arising out of the use of TI products in
such safety-critical applications.
TI products are neither designed nor intended for use in military/aerospace applications or environments unless the TI products are
specifically designated by TI as military-grade or "enhanced plastic." Only products designated by TI as military-grade meet military
specifications. Buyers acknowledge and agree that any such use of TI products which TI has not designated as military-grade is solely at
the Buyer's risk, and that they are solely responsible for compliance with all legal and regulatory requirements in connection with such use.
TI products are neither designed nor intended for use in automotive applications or environments unless the specific TI products are
designated by TI as compliant with ISO/TS 16949 requirements. Buyers acknowledge and agree that, if they use any non-designated
products in automotive applications, TI will not be responsible for any failure to meet such requirements.
Following are URLs where you can obtain information on other Texas Instruments products and application solutions:
Products
Applications
Audio
www.ti.com/audio
Communications and Telecom www.ti.com/communications
Amplifiers
amplifier.ti.com
Computers and Peripherals
www.ti.com/computers
Data Converters
dataconverter.ti.com
Consumer Electronics
www.ti.com/consumer-apps
DLP® Products
www.dlp.com
Energy and Lighting
www.ti.com/energy
DSP
dsp.ti.com
Industrial
www.ti.com/industrial
Clocks and Timers
www.ti.com/clocks
Medical
www.ti.com/medical
Interface
interface.ti.com
Security
www.ti.com/security
Logic
logic.ti.com
Space, Avionics and Defense
www.ti.com/space-avionics-defense
Power Mgmt
power.ti.com
Transportation and
Automotive
www.ti.com/automotive
Microcontrollers
microcontroller.ti.com
Video and Imaging
www.ti.com/video
RFID
www.ti-rfid.com
Wireless
www.ti.com/wireless-apps
RF/IF and ZigBee® Solutions
www.ti.com/lprf
TI E2E Community Home Page
e2e.ti.com
Mailing Address: Texas Instruments, Post Office Box 655303, Dallas, Texas 75265
Copyright © 2011, Texas Instruments Incorporated
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising