Sains Malaysiana 39(2)(2010): 291–297

Pengecaman Aksara Jawi Menggunakan Jelmaan Surih

(Jawi Character Recognition using the Trace Transform)

Mohammad Faidzul Nasrudin, Khairuddin Omar & Mohamad Shanudin Zakaria

Center for Artificial Intelligence Technology (CAIT)

Faculty of Information Sciences and Technology, Universiti Kebangsaan Malaysia

43600 UKM Bangi, Selangor D.E., Malaysia

Choong-Yeun Liong*

Pusat Pemodelan dan Analisis Data (DELTA), Pusat Pengajian Sains Matematik

Fakulti Sains dan Teknologi, Universiti Kebangsaan Malaysia

43600 UKM Bangi, Selangor D.E., Malaysia

Received: 17 April 2009 / Accepted: 16 September 2009

ABSTRAK

Jelmaan surih yang merupakan pengitlakan jelmaan Radon, membenarkan pembinaan fitur imej tak-ubah kepada sekumpulan jelmaan imej yang dipilih. Dalam makalah ini, penulis mendemonstrasi kebergunaan fitur Jelmaan surih yang tak-ubah kepada herotan afin bagi membolehkannya membezakan aksara Jawi. Proses ini terdiri daripada menyurih imej dengan garis-garis lurus pada semua orientasi yang mungkin sambil menghitung beberapa fungsian bagi fungsi imej. Setiap kombinasi fungsian akan menghasilkan satu fungsi orientasi (atau fitur) bagi garis-garis surihan tersebut yang dikenali sebagai tandatangan objek. Jika fungsian yang digunakan mempunyai beberapa sifat pratakrif, tandatangan objek tersebut boleh digunakan untuk membezakan aksara Jawi secara afin. Ia bermanfaat untuk membina fitur tak-ubah terhadap putaran, translasi, penskalaan dan ricihan imej. Seterusnya, penulis mendemonstrasi kebergunaan fitur ini dengan membandingkan keputusan pengecamannya dengan keputusan yang diperoleh daripada fitur berasaskan momen afin tak-ubah. Eksperimen menggunakan Jelmaan surih telah menghasilkan keputusan yang cemerlang untuk pengecaman aksara Jawi bercetak dan tulisan tangan yang tak-ubah kepada herotan afin.

Kata kunci: Jelmaan surih; momen afin tak-ubah; pengecaman aksara Jawi

ABSTRACT

The Trace transform, a generalisation of the Radon transform, allows one to construct image features that are invariant to a chosen group of image transformations. In this paper, we used some features, which are invariant to affine distortion, generated by the Trace transform to discriminate between Jawi characters. The process consists of tracing an image with straight lines, along which certain functionals of the image function are calculated, in all possible orientations. For each combination of functionals we derived a function of orientation of the tracing lines that is known as an object signature. If the functionals used have some predefined properties, this signature can be used to characterise the character in an affine way. We demonstrated the usefulness of the derived signature and compared the result of character recognition with those obtained by using features based on affine moment invariants. Experiments using the Trace transform produced decent results for the printed and handwritten Jawi character recognitions that are invariant to affine distortion.Keyword: Affine moment invariant; Jawi character recognition; trace transform

REFERENCES

Baghshah, M.S. 2005. A Novel Fuzzy Approach to Recognition of Online Persian Handwriting. Proceedings of the 5th International Conference on Intelligent Systems Design and Applications 2005 (ISDA ‘05), 8-10 September.Wroclaw, Poland: 268-273.

Deans, S.R. 1983. The Radon Transform and Some of Its Applications. Berlin: Krieger.

Flusser, J. & Suk, T. 1993. Pattern Recognition by Affine Moment Invariants. Pattern Recognition 26: 167-174.

Flusser, J. & Suk, T. 1994a. Affine Moment Invariants: A New Tool for Character Recognition. Pattern Recognition Letters 15: 433-436.

Flusser, J. & Suk, T. 1994b. A Moment-Based Approach to Registration of Images with Affine Geometric Distortion. IEEE Transactions on Geoscience and Remote Sensing 32: 382-387.

Kadyrov, A. & Fedotov, N. 1995. Triple Features Pattern Recognition and Image Analysis. Advances in Mathematical Theory and Applications 5(4): 546-556.

Kadyrov, A. & Petrou, M. 1998. The Trace Transform as a Tool to Invariant Feature Construction. Proceedings of ICPR’98, Brisbane, Australia: 1037-1039.

Kadyrov, A. & Petrou, M. 2001a. Object Descriptors Invariant to Affine Distortions. Proceedings of BMVC 2001, Manchester, UK 2: 391-400.

Kadyrov, A. & Petrou, M. 2001b. The Trace Transform and Its Applications. IEEE Transactions on Pattern Analysis and Machine Intelligence 23(8): 811-828.

Kadyrov, A., Petrou, M. & Park, J. 2001. Korean Character Recognition with the Trace Transform. Proceedings of the International Conference on Integration of Multimedia Contents (ICIM 2001), November 15, Chosun University, Gwangju, South Korea: 7-12.

Kadyrov, A. & Petrou, M. 2003. Object Signatures Invariant to Affine Distortions Derived from the Trace Transform. Image and Vision Computing (IVC) 21(13-14): 1135-1143.

Khairuddin Omar. 2000. Pengecaman Teks Tulisan Tangan Jawi Menggunakan Pengelas Multi-Aras. Tesis Ph.D. Universiti Putra Malaysia.

Lorigo, L.M. & Govindaraju, V. 2006. Offline Arabic Handwriting Recognition: A Survey. IEEE Transactions on Pattern Analysis and Machine Intelligence 28(5): 712-724.

Mazani Manaf 2002. Pengecaman Teks Tulisan Tangan Jawi Menggunakan Rangkaian Neural Recurrent Bama. Tesis Ph.D. Universiti Kebangsaan Malaysia.

Mohammad Faidzul Nasrudin, Khairuddin Omar, Mohamad Shanudin Zakaria & Liong Choong-Yeun 2008a. Handwritten Cursive Jawi Character Recognition: A Survey. Proceedings of the 5th International Conference on Computer Graphics, Imaging and Visualization: 247-256.

Mohammad Faidzul Nasrudin, Liong Choong-Yeun Khairuddin Omar & Mohamad Shanudin Zakaria 2008b. Trace transform for Jawi Character Recognition. Dlm. Proceedings of the International Symposium on New Development of Geometric Function Theory and its Applications (GFTA 2008), edited by M. Darus & S. Owa. School of Mathematical Sciences, Universiti Kebangsaan Malaysia: 441-448.

Petrou, M. & Kadyrov, A. 2004. Affine Invariant Features from the Trace Transform. IEEE Transactions on Pattern Analysis and Machine Intelligence 26(1): 30-44.

Plamondon, R. & Srihari, S.N. 2000. On-Line and Off-Line Handwriting Recognition: A Comprehensive Survey. IEEE Transactions on Pattern Analysis and Machine Intelligence 22(1): 63-84.

Radon, J. 1917. Uber die Bestimmung vonFunktionen durch ihre Integralwerte langesgewisser Mannigfaltigkeiten. Ber. SachsischeAkademie der Wissenschaften 69: 262-277.

Shin, B.S., Cha, E.Y., Cho, K.W., Klette, R. & Woo, Y.W. 2008. Effective Feature Extraction by Trace Transform for Insect Footprint Recognition. Multimedia Imaging Report 12. Computer Science Department: The University of Auckland, New Zealand.

*Corresponding author; email: lg@ukm.my