B ezier Curve Data Reconstruction over Compressive Sampling in Image Transmission

Syamsul Rizal, and Dong-Seong Kim School of Electronic Engineering Kumoh National Institute of Technology, South Korea e-mail: syamsulrizal123@gmail.com, dskim@kumoh.ac.kr
AbstractThis paper proposed a method for image reconstruction using Bezier Curve function. This method compared with the previous works, compressive sampling and discrete Fourier transform (DFT). Compressive sampling is a method which has small number of data sample and DFT used to perform Fourier analysis in many practical applications in digital signal processing. One of the causes of errors in image reconstruction is lack of sampling. In order to solve that problem, this paper proposed a method by using Bezier Curve to interpolate data between sampling points to reduce the error. The result shows that image reconstruction using Bezier curve has close to the original image compared to compressive sampling and DFT. Index TermsB ezier curve, compressive sampling, DFT, image reconstruction.

Fig. 1. The data frame structure: a) Compressive data transmission, b) Functional data transmission

I. I NTRODUCTION Image or video processing is any form of signal processing where the input is an image, such as photographs or video frame, while the output of image processing can be either an image or a set of characteristics or parameters related to image. Many techniques can be used in image processing such as compressive sampling, discrete Fourier transform (DFT), etc. Compressive sampling or compressive sensing (CS) is a signal processing method for efciently and reconstruction a signal but contradicted with the common method in data acquisition. This technique can reduces power consumption by requiring acquisition and transmission of fewer samples [1]. The signal can be reconstructed by using fewer sampling compared to traditional method such as Nyquist theorem, where the sampling data must be at least twice the maximum frequency used in the signal [2], [3]. Besides compressive sampling, Discrete Fourier Transform also a technique used for image processing. DFT is used to transform original image from spatial to frequency domain [4], [5]. In image processing, the samples can be the values of pixels along a row or column of an image. This paper proposed a technique to reconstructed the image using B ezier curve function to interpolate the data between each sampling point. B ezier curve is a set of independent variables that represent the coordinates of several points in curved line between two points. Hence, the data between sampling points can be predicted and reconstructed using this method. Using image processing technique, the image size will become smaller than the original image with good quality. These techniques can increases energy efcient and send the

Fig. 2. Cubic B ezier curve.

image faster than the original due to the small size of the image [2]. The remain of this paper is organized as follows: Section II describes the compressive sampling and DFT, section III explains the proposed method. Results and analysis are explained in section IV. Finally, section V concludes the paper. II. I MAGE P ROCESSSING T ECHNIQUES Image processing is a technique to compressed and reconstructed the image to become the image size was smaller than the original image but with the image that resembles the actual image. Many techniques can be used to reconstructed image in image processing, such as compressive sampling, DFT, etc. A. Compressive Sampling Overview Compressive sampling is a compression method with a random collection an amount of data which is determined by the value of the rate measurement. This method used 2 transformations; sparsity and incoherent transform [2], [3]. A sensing mechanism in which information about signal f (t) is obtained by linear functional recording this value. yk = f, k k = 1, ..., m. (1)

That correlates the object that we wish to acquire with the waveforms k (t). y is a vector of sampled values of f in

the time or space domain, Dirac delta functions (spikes) are sensing waveform. If the sensing waveforms are indicator functions of pixels, then y is the image data typically collected by sensors in a digital camera. If the sensing waveforms are sinusoids, then y is a vector of Fourier coefcients. 1) Sparsity: Sparsity expresses the idea that the information rate of a continuous signal may be much smaller than suggested by its bandwidth. Mathematically, a vector f Rn (such as the n-data) which expanded in orthonormal basis (such as a wavelet basis) = [1 , 2 , ..., n ] as follows:
n

f (t) =
i=1

xi i (t),

(2)

where x is the coefcient sequence of f, xi = (f, i ). It will be convenient to express f as x , where is the nxn matrix with 1 , ..., n as columns. The implication of sparsity is now clear: when a signal has a sparse expansion, one can discard the small coefcients without much perceptual loss. 2) Incoherence: Incoherence extend the duality between time and frequency and expressed the idea that objects having a sparse representation in must be spread out in the domain in which they are required. Suppose there is a pair (, ) of orthobases of Rn . The rst basis is used for sensing the object f as in (1) and the second is used to represent f . The coherence between the sensing basis and the representation basis is (3) (, ) = n. max | k , j |.
1k,j n

Fig. 3. B ezier curve transmission process

The coherence measures the target correlation between any two elements of and . If and contain correlated elements, the coherence is large. Otherwise, it is small. As for how large or small, it follows from linear algebra that (, ) [1, n]. B. Fourier Transform Overview In mathematics, the discrete Fourier transform (DFT) is one of the spesic form of Fourier analysis. It transforms one function to another, which is called the frequency domain representation of the time domain. Yet the DFT requires an input function that is discrete and whose non-zero values have a limited (f inite) duration. Since the input function is a nite sequence of real or complex numbers, the DFT is ideal for processing information stored in computers. In particular, the DFT is widely employed in signal processing and related elds to analyze the frequencies contained in a sampled signal, to solve partial differential equations, and to perform other operations such as convolutions. The DFT can be computed efciently in practice using a fast Fourier transform (FFT) algorithm. FFT algorithms are so commonly employed to compute the DFT, the two terms are often used interchangeably in colloquial settings, although there is a cleal distinction: DFT refers to a mathematical transformation, regardless of how it is computed, while FFT refers to any one of several efcient algorithms for the DFT.

The sequence of N complex numbers x0 ,...,xN 1 is transformed into the sequence on N complex numbers X0 ,...,XN 1 by the DFT according to the formula: where is a primitive N th root of unity. III. S YSTEM M ODEL Data frame structure is a basic method to describe this proposed method. Figure 1 shows compressive sampling and proposed method data frame structure. The data frame is consisted of data and protector part. Protector part usually represented as the error correction such as CRC, ECC, etc. This paper proposed a progressive data transmission using B` ezier curve formula, which is described in Figure 1. A. B ezier Curve B` ezier curve is a set of independent variables that represent the coordinates of several points in curved line between two or more points. A B ezier curve is dened by a set of control points P0 through Pn , where n is B eziers order (n=1 for linear, 2 for quadratic, etc). The rst and last control points are always the end points of the curve. However, the intermediate control points generally do not lie on the curve. Consider the following cubic-B ezier formulation. Dene A is the set of number where a0 = |A|min and an1 = |A|max . Therefore, there exist such a P that represents A where |P | |A| and A C . Consider the cubic-B ezier where the curve is represented by 4 points. Figure 2 shows the possibilities to cover a data

a b c d
Fig. 4. image reconstruction : a. Original, b. DFT, c. Compressive Sampling, d. B ezier Curve

which consists of 5 numbers data that lies on the curve within a set that can be represented 4 numbers of cubic-B ezier. It means that by sending 4 data from B ezier function, we can construct 5 data on the receiver. The extended version of the cubic-B ezier, double cubic B ezier, where 6 numbers of B ezier function, can be used to represent 9 numbers of data. This process can be described as the subdividing process of the curve. In other words, every curve can be scooped within the B ezier function depends on the depth of the curve. Therefore, the points of one B ezier curve can be extended as follows,
j 1 1 Pj + Pj i = (1 )Pi i+1 .

Algorithm 1 Constructing B ezier Curve t is ratio between 0-1 for an order n curve do Take all lines between the curve dening points Place markrers along each of these line, at distance t if t=0.2 then Place mark at 20% from the start end if Form lines between those points, this gives n-l lines Repeat this until have only one line left end for

(4) B ezier Curve can be constructed using Algorithm 1. Treat t as a ratio, t=0 is 0% along a line t=100% along a line. Take all lines between the curves dening points. For an order n line, that is n lines. Place markers along each of these line, at distance t. So if t=0.2, place a mark at 20% from the start, 80% from the end. Now form lines between those points. This gives

Eq.(4) shows that every data can be approached by any B ezier curved. The detail level is formulated corresponds directly with the level of B ezier curve division. The transmission process of B ezier curve can be seen at Figure 3. If the errors occurs during transmission, the B ezier curve construction is performed by the following procedure.

n-l lines. Place markers along each of these line at distance t. Form lines between those points, this will be n-2 lines. Repeat this until have only one line left. The point of t on that line coincides with the original curve point at t. B. Proposed Method Starting from the S received data, the receiver creates sampling data by using compressive sampling in P number of sampling which is much less than S. For every 4 P points of sampling data (P0 ,P1 ,P2 ,P3 ), B ezier curve is applied to interpolate the data. Let P0 and P3 are the start and the end point of B ezier curve, P1 and P2 are intermediate points which are used as control points to create the B ezier curve, Fig. 2. This algorithm is repeated until the B ezier curve covers all P sampling points. After B ezier curve is obtained, the receiver extracts its function and reconstructs the data based on B ezier curve. IV. R ESULTS AND A NALYSIS In this simulation, we consider image 100x100 pixels to be reconstructed. We compare the reconstructed image by using compressive sampling, DFT, and B ezier curve. We use standard DFT and 10 percent of data sampling in compressive sampling. Figure 4 shows 4 simulation results of image reconstruction. Figures 4a are the original image which would be reconstructed. In both of the Figures 4b and 4c show the image cannot be reconstructed well. The quality of the images are poor. In Figures 4d show the quality of image are better than the other method (CS and DFT). With the proposed method the image reconstructed clear and well. Beside the image quality, compressed technique is useful for reduce the size of image. Table 1 shows the size different from each method. From that Table we can see the image size using proposed method is smaller than compressive sampling and DFT methods. Using this method we can reduce the image size become smaller but with good image quality. The size of an image will affect the transmission time in the network. Figure 5 shows the transmission time for every method. In assumption, the bandwidth in the network is 10KBps. The graph shows that the image which has been reconstructed using B ezier curve is faster than the other methods. It means we can save time to transfer image from the source to the destination with good image quality. V. C ONCLUSION In this paper, an approach to reconstruct data by taking advantage of B ezier curve in compressive sampling and DFT was presented. Compressive sampling is a sampling method that against sampling paradigm that the number of sampling is much less than the original data. This method caused data error in reconstruction. DFT is a method which is used commonly in image processing. This paper presented a method to help image reconstruction by adding a function called B ezier curve to interpolate data between sampling points the reconstruct data. The cubic B ezier curve is used in this paper, this method takes 4 sampling points to interpolate and construct

No. 1

Image Size (KByte) Original a1 24 Original a2 14 Original a3 19 Original a4 23 DFT b1 24 DFT b2 18 DFT b3 20 DFT b4 18 CS c1 10 CS c2 32 CS c3 32 CS c4 31 B ezier Curve d1 3 B ezier Curve d2 2 B ezier Curve d3 3 B ezier Curve d4 3 TABLE I I MAGE S IZE

Fig. 5. Transmission time of images

the image (100x100 pixels). By using this method, image reconstruction is closer to the original data compared to only use the compressive sampling or DFT methods. For future study, we will develop a method to suppress the constructed data to be same as original data. Another way is by adding error correction in data it self to help data reconstruction. R EFERENCES
[1] G. Orchard, J. Zhang, Y. Suo, M. Dao, D. Nguyen, S. Chin, C. Posch, T. Tran, and R. Etienne-Cummings, Real Time Compressive Sensing Video Reconstruction in Hardware, IEEE Journal on Emerging and Selected Topics in Circuits and Systems, vol. 2, no. 3, pp. 604 615, September 2012. [2] F. Chen, F. Lim, O. Abari, A. Chandrakasan, and V. Stojanovic, EnergyAware Design of Compressed Sensing Systems for Wireless Sensors Under Performance and Reliability Constraints, IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 60, no. 3, pp. 650 661, March 2013. [3] X. Wang, Z. Zhao, N. Zhao, and H. Zhang, On the application of compressed sensing in communication networks, 5th International ICST Conference on Communications and Networking in China (CHINACOM), pp. 1 7, August 2010. [4] J. Mandal and A. Khamrui, A Genetic Algorithm based steganography in frequency domain (GASFD), International Conference on Communication and Industrial Application (ICCIA), pp. 1 4, December 2011. [5] B. Varma, K. Paul, and M. Balakrishnan, 26th international conference on vlsi design and 12th international conference on embedded systems (vlsid), Accelerating 3D-FFT Using Hard Embedded Blocks in FPGAs, pp. 92 97, January 2013.