The Processing Time Comparison

The proposed system adopts the CAM for the storage due to its better performance than the general storage RAM, and the features of a CAM have been introduced in chapter 2. In this section, the proposed CAM-based LPR system with the improved PAV method and a general RAM-based LPR system with the same method will both be tested in order to discover the processing time differences. To keep the same experimental environments, the clock frequencies are both set as 25MHz; both systems are established and simulated on the same development board, EP20K1500EBC652-1X, and of course, both systems use the same database for the experiment.

The analysis is based on the processing time result of each test input image in the database. With the total 1800 test images, all the processing time corresponding to the proposed CAM-based system and the general RAM-based system is shown in Figure 4-3, and the average processing time of the two systems is listed in Table 4-7. From the statistical information, it obviously shows that the proposed CAM-based system can save more processing time consumption about 50.4μs than the RAM-based system.

Figure 4-3: The processing time between CAM and RAM.

Table 4-7: The average processing time of CAM-based and RAM-based systems.

average processing time

CAM-based 77.998μs

RAM-based 128.367μs

Chapter 5

Conclusions

License plate character recognition system has become the key to many traffic related applications such as the traffic enforcement systems and the electronic toll-collection systems. However, most of the developed license plate recognition systems are PC-based due to the use of complicated algorithms.

This thesis has implemented the license plate character recognition system on the DSP board (SN: EP20K1500EBC652-1X) to verify the potential of a hardware system other than PC-based. The system adopts a specific storage called the Content Addressable Memories to replace the common RAM and further develops the module of a 9-bits 512-words CAM. A CAM is better than a RAM due to the unique architecture and the parallel searching function. An image binarization test has shown the capability of the CAM for the image processing application.

For the algorithm of the character recognition, the PAV method with a unique feature vector called the Pattern Accumulated Vector has been developed. The PAV method selects the principal pattern blocks in an efficient way and recognizes the

images by less complicated operations. Furthermore, the basic PAV method and the improved PAV method have been derived for the normal templates and the deformed templates, respectively.

Through the simulations and experiments, the recognition reaches a rate of 99.78% with the use of the improved PAV method. The failures in the characters are mostly due to the extreme cases of inclined, fragmental, and noised conditions. On the other hand, the average processing time of the proposed system is 77.998μs, which is about 50.4μs faster than the RAM-based system. These results confirm that the proposed system has potential and is feasible.

A complete LPR system includes the license plate extraction and the character segmentation, which are both needed for the proposed system to upgrade. In addition, a complete hardware containing the input interface, the DSP chip, and the output display, will be ready in the future.

Reference

[1] Y. Amit, D. Geman, and X. Fan, “A Coarse-to-Fine Strategy for Multiclass Shape Detection,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 26, no. 12, pp. 1606-1621, Dec. 2004.

[2] H. Anderson, “Neural network machines,” Potentials, IEEE, vol. 8, no. 1, pp.

13-16, Feb. 1989.

[3] R. F. Chang and W. M. Chen, “Adaptive Edge-based Side-match Finite-state Classified Vector Quantization with Quadtree Map,” IEEE Transactions on Image Processing, vol. 5, no. 2, pp. 378-383, Feb. 1996.

[4] S. L. Chang, L. S. Chen, Y. C. Chung, and S. W. Chen, “Automatic License Plate Recognition,” IEEE Transactions on Intelligent Transportation Systems, vol. 5, no. 1, pp. 42-53, Mar. 2004.

[5] L. Chisvin and R. Duckworth, “Content-Addressable and Associative Memory:

Alternatives to the ubiquitous RAM,” IEEE Transactions on Computer, vol. 22,

no. 7, pp. 51-64, Jul. 1989.

[6] C. Foster, Content Addressable Parallel Processors, John Wiley & Sons, Inc., New York. USA, 1976.

[7] T. Ikenaa and T. Ogura, “A Fully Parallel 1-MB CAM LSI for Real-Time Pixel-Parallel Image Processing,” IEEE Journal of Solid-State Circuits, vol. 35, no. 4, pp. 536-544, Apr. 2000.

[8] S. Impedovo, L. Ottaviano, S. cchinegro, “Optical Character Recognition - A Survey,” International Journal of Pattern Recognition and Artificial Intelligence, vol. 5, no. 1-2, pp. 1-24, 1991.

[9] T. Jamil, “RAM versus CAM,” Potentials, IEEE, vol. 16, no. 2, pp. 26-29, 1997.

[10] S. Kim, D. Kim, Y. Ryu, and G. Kim, “A Robust License Plate Extraction Method Under Complex Image Conditions,” in Proc. 16th International Conference on Pattern Recognition (ICPR’02), Quebec City, Canada, vol. 3, pp.

216-219, Aug. 2002.

[11] S. Kornachuk, L. McNaughton, R. Ghibbins, and N. D. Benoit, “A High Speed Embedded Cache Design With Non-intrusive BIST,” in Proc. IEEE 1994 MTDT Conference, San Jose, Canada, pp. 40-45, Aug. 1994.

[12] A. Krikelis and C. Weems, “Associative Processing and Processors,” IEEE Transactions on Computers, vol. 27, no. 11, pp. 12-17, Nov. 1994.

[13] C. S. Lin, J. C. Chang, and B. D. Liu, “A Low-power Pre-computation Based Fully Parallel Content Addressable Memory,” IEEE Journal of Solid-State Circuits, vol. 38, no. 4, pp. 654-662, Apr. 2003.

[14] K. J. Lin, C. W. Wu, and S. Member, “A Low-power CAM Design for LZ Data Compression,” IEEE Transactions on Computers, vol. 49, no. 10, pp. 1139-1145, Oct. 2000.

[15] V. Madisetti, VLSI Digital Signal Processors: An Introduction to Rapid Prototyping and Design Synthesis, IEEE Press, 1995.

[16] I. Masaki, “Machine-Vision Systems for Intelligent Transportation Systems,”

IEEE Intelligent Systems, vol. 13, no. 6, pp. 24-31, Nov. 1998.

[17] B. McQueen and J. McQueen, Intelligent transportation systems architectures, Artech House, Boston. USA, 1999.

[18] M. Milward, N. Nunez, and D. Mulvaney, “Design and Implementation of a Lossless Parallel High-Speed Data Compression System,” IEEE Transactions on Parallel and Distributed Systems, vol. 15, no. 6, pp. 481-490, Jun. 2004.

[19] M. Motomura, J. Toyoura, K. Hirata, H. Ooka, H. Yamada, and T. Enomoto, “A 1.2-million transistor 33-MHz 20-b dictionary search processor (DISP) ULSI with a 160-kb CAM,” IEEE Journal of Solid-State Circuits, vol. 25, no. 5, pp.

1158-1165, Oct. 1990.

[20] T. Naito, T. Tsukada, K. Yamada, K. Kozuka, and S. Yamamoto, “Robust License-plate Recognition Method for Passing Vehicles under Outside Environment,” IEEE Transactions on Vehicular Technology, vol. 49, no. 6, pp.

2309-2319, Nov. 2000.

[21] K. Nogami, T. Sakurai, K. Sawada, K. Sakaue, Y. Miyazawa, S. Tanaka, Y.

Hiruta, K. Katoh, T. Takayanagi, T. Shirotori, Y. Itoh, M. Uchida, and T. Iizuka,

“A 9-ns Hit-delay 32-kbyte Cache Macro for High-speed RISC,” IEEE Journal of Solid-State Circuits, vol. 25, no. 1, pp. 100-106, Feb. 1990.

[22] B. Ramamurthi and A. Gersho, “Clasified vector quantization of images,” IEEE Transactions on Communications, vol. COM-34, no. 11, pp. 1105–1115, Nov.

1986.

[23] S. H. Seo and M. R. Azimi-Sadjadi, “A 2-D Filtering Scheme for Stereo Image Compression Using Sequential Orthogonal Subspace Updating,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 11, no. 1, pp.

52-66, Jan. 2001.

[24] V. Shapiro and G. Gluhchev, “Multinational License Plate Recognition System:

Segmentation and Classification,” in Proc. 17th International Conference on Pattern Recognition (ICPR’04), Cambridge, UK, pp. 352-355, Aug. 2004.

[25] K. W. Wong, K. M. Lam, and W. C. Siu, “An Efficient Low Bit-Rate Video-Coding Algorithm Focusing on Moving Regions,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 11, no. 10, pp. 1128-1134, Oct.

2001.

[26] H. Yamada, Y. Murata, T. Maeda, R. Ikeda, K. Motohashi, and K. Takahashi,

“Real-time String Search Engine LSI for 800-Mbit/s LANs,” in Proc. IEEE 1988 Custom Integrated Circuits Conference., Rochester, USA, pp. 21.6.1-21.6.4,

May 1988.

[27] B. Yan and K. W. Ng, “A Novel Selective Motion Vector Matching Algorithm for Error Concealment in MPEG-4 Video Transmission over Error-Prone Channels,”

IEEE Transactions on Consumer Electronics, vol. 49, no. 4, pp. 1416-1423, Nov.

2003.