International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459,ISO 9001:2008 Certified Journal, Volume 2, Issue 12, December 2012)
667
Motion Estimation and Motion Compensated Video
Compression Using DCT And DWT
Thazni Aziz
1, D. Raveena Judie Dolly
21PG Scholar, 2Assistant Professor, Dept. of ECE, Karunya University,Coimbatore – INDIA
Abstract— Video compression plays an important role in real-time scouting/video conferencing applications. For the entire motion based video compression process, motion estimation is the most computationally expensive and time-consuming process. Block matching techniques are the most popular and efficient of the various motion estimation techniques. Block matching algorithm in video compression select the current frame and divides in to blocks. These, helps to find motion vector for each blocks within a search range find a best match that minimize an error measure. In this paper, Full Search strategies are used to reduce computation. Video compression techniques are used to reduce the redundancy in video data. For this purpose DCT (Discrete Cosine transform) and DWT (Discrete Wavelet Transform) are used. One advantage of DCT, Find the match of low frequency values then it can increase into comparing the higher frequency. The goal of wavelet based compression is to store video data in a little space. Hence, analyzed the performance is analyzed based on compression ratio and PSNR values using these two techniques.
Keywords— compression, DCT, DWT, motion
estimation, motion vector.
I. INTRODUCTION
Video compression techniques are used to reduce redundancy in video data without affecting visual quality. It mostly used in video conference and real time application[4]. In reality, motion estimation based encoders are the most widely used in video compression techniques. Such encoders make use of inter-frame correlation to provide well-organized compression [3].3D transform coder produces video compression ratio which is close to the motion estimation based coding one with less complex processing. Redundancy has not the same pertinence since the efficiency of 3D transform can reduce as pixel’s values variation in spatial or temporal dimensions is not uniform.
Often the temporal redundancies are more relevant than spatial one [3].
Methods for compression There are four methods for compression
1. Discrete Cosine Transform (DCT)
2. Discrete Wavelet Transform (DWT)
Motion compensation technique engaged in the encoding of video data for video compression. Motion compensation describes the transformation of a reference picture to the current picture [6]. The reference picture may be previous or even taken the later frames. When current images can be accurately synthesized from previously transmitted or stored images[5], the compression efficiency can be improved.
In this paper DCT (Discrete Cosine transform) and DWT (Discrete Wavelet Transform) are used for video compression. Block matching algorithm, helps to find motion vector for each blocks within a search range find a best match that minimize an error measure.
II. METHODOLOGY
Now a days, the communication area faces a big challenge because of large size videos. So video compression technique is save storage space. These technique find the redundancies in the move frame and the correlation between the scenes [7]. The proposed approach is based on the DCT/DWT approach to perform the video
compression with scalability vector. Here DWT will
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459,ISO 9001:2008 Certified Journal, Volume 2, Issue 12, December 2012)
668
Fig.1. Block diagram of video compression
A. Motion Estimation
For the entire motion based video compression process, motion estimation is the most expensive and time-consuming process. Block matching techniques are the most popular and efficient methods for various motion estimation techniques. For the reference frame, a search area is defined for each block in the current frame. The search area is typically sized at 2 to 3 times the macroblock size (16x16) [4]. Using the fact that the motion between consecutive frames is statistically small, the search range is confined to this area. After the search process, a best match will be found within the area. The best matching usually means having lowest energy in the sum of residual formed by subtracting the candidate block in search region from the current block located in current frame. The process of finding best match block by block is called block-based motion estimation. This can be used to find the motion vectors in a current frame.
B. Motion Compensation
Motion compensation is an algorithmic technique employed in the encoding of video data for video compression. The only difference between two frames is [4], either the camera moving or an object in the frame moving. Mean Absolute Difference (MAD) is used for finding the difference of two frames. The MAD equation is given below,
MAD = (1)
Where N is the side of the macrobock, Cij and Rij are the pixels being compared in current macroblock and reference macroblock. In a video file much of the information that represents one frame will be the same as the information used in the next frame.
For finding the similarity between two frames, threshold value 0.06 was set.
C. DCT (Discrete Cosine Transform)
A discrete cosine transform (DCT) express a sequence of many data points in terms of a sum of cosine functions at
different frequencies. DCTs are important to lossy
compression of audio (e.g. MP3) and images (e.g. JPEG). Disadvantages of DCT
Only spatial correlation of the pixels inside the single 2-D block is considered and the correlation from the pixels of the neighboring blocks is neglected
Impossible to completely decorrelate the blocks at their boundaries using DCT
Undesirable blocking artifacts affect the reconstructed images or video frames. (high compression ratios or very low bit rates.
Poor identification of which data is relevant to human perception less compression ratio
D. DWT (Discrete Wavelet Transform)
The Discrete Wavelet Transform passing a signal to image, through a pair of filters, a low pass filter and a high pass filter. The low pass filter yields low resolution signal. The high pass filter yields difference signal. The outputs of these filters are downsampled by two. The downsampled outputs have the same number of bits as the input signal. The original signal is reproduced, when the upsampled output of the low pass filter is added to the upsampled output of the high pass filter. The output of the high pass filter is fed into another pair of filters and the process repeated. Haar wavelet transform is the simple example of discrete wavelet transforms [4]. The wavelet transform (WT) has gained widespread acceptance in signal processing and image compression. Because of their inherent multi-resolution nature, wavelet-coding schemes are especially suitable for applications where scalability and tolerable degradation are important. Recently the JPEG committee has released its new image coding standard, JPEG-2000, which has been based upon DWT. Discrete wavelet transform (DWT), which transforms a discrete time signal to a discrete wavelet representation.
III. RESULT AND DISCUSSION
A. Input Video Sequence
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459,ISO 9001:2008 Certified Journal, Volume 2, Issue 12, December 2012)
669
Rhinos.avi file is used for a video compression. First, uncompressed videos are converted into many frames.fig 7 shows the video frames.
(a) (b) (c)
Fig.2. Uncompressed video sequence (a),(b),(c).
B. Motion Estimation
Motion estimation is the process of determining motion vectors that describe the transformation from one 2D image to another; usually from adjacent frames in a video sequence. The motion vectors may relate to the whole image (global motion estimation) or specific parts, such as rectangular blocks, arbitrary shaped patches or even per pixel.
Fig.3. Motion vector
The motion vectors may be represented by a translational model or many other models that can approximate the motion of a real video camera, such as rotation and translation in all three dimensions and zoom.
C. Motion Compensation
In Motion Compensation selected two frames such as current frame and reference frame. The selected frames are converted into binary format and find the difference between these two frames using Mean absolute difference.
(a) Frame #30 (b) Frame#36
Fig.4. Motion compensated Reference frames (a and b)
Fig.5. Absolute Difference with motion compensation
(frame #30, frame# 36)
D. Discrete Cosine Transform (DCT)
Here the input is the residue picture calculated by the motion estimation unit. Since the residue picture has high correlation between neighboring pixels, the transformed data is easier to compress than the original residue data since the energy of the transformed data is localized. The transformed data are called transform coefficients and they are passed to the quantization unit. The transformation can be done by many methods, including the Cosine Transform.
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459,ISO 9001:2008 Certified Journal, Volume 2, Issue 12, December 2012)
670
(c)
Fig.6. Images used to evaluate the DCT performance.(a) Residue image,(b) Predicted image,(c) Compressed image,
CR=0.5763,PSNR=51.546
E. Discrete Wavelet Transform (DWT)
Wavelet transform decomposes a signal into a set of basis functions.These basis functions are called wavelets. Here, three types of decomposing levels are used. The difference of compression ratio between these three levels
is shown below. The wavelet transform is computed
separately for different segments of the time-domain signal at different frequencies. Multi-resolution analysis: analyzes the signal at different frequencies giving different resolutions
Fig.7. DWT image
1).1 level decomposition
(a) (b)
Fig.8. Image used to evaluate1 level decomposition.(a) input image,(b) compressed image CR=91.2813, PSNR= 63.897
2). 2 Level Decomposition
(a) (b)
Fig.9. Image used to evaluate 2 level decomposition. (a) Input image,(b) compressed image.CR=88.1616, PSNR=58.063
3). 3 Level Decomposition
(a) (b)
Fig.10. Image used to evaluate 3level decomposition.(a) Input image, ( b) Compressed image CR=83.72,PSNR=53.897
Advantages of DWT over DCT
1. No need to divide the input coding into
non-overlapping 2-D blocks, it has higher compression ratios avoid blocking artifacts.
2. Allows good localization both in time and spatial frequency domain.
3. Transformation of the whole image introduces
inherent scaling
4. Better identification of which data is relevant to human perception higher compression ratio. 5. Higher flexibility: Wavelet function can be freely
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459,ISO 9001:2008 Certified Journal, Volume 2, Issue 12, December 2012)
[image:5.612.82.256.154.306.2]671
TABLE I
COMPARISSON BETWEEN DCT and DWT
IV. CONCLUSION
Video compression techniques are used to reduce the redundancy in video data. For this purpose DCT (Discrete Cosine transform) and DWT (Discrete Wavelet Transform) are used. Block matching algorithm, helps to find motion vector for each blocks within a search range and finds a best match that minimize an error measure. the performance of DCT and DWT technique is compared based on compression ratio was found.That DWT performance better than DCT.
REFERENCES
[1 ] Jaya Krishna Sunkara,D.Pradeep,D.Pavani. 2012 A new video compression method using DCT/DWT and SPHIT based on accordion representation.
[2 ] Jasmeetkanur,Rohini Sharma. 2012 A combined DWT-DCT approach to perform video compression base of frame redundancy.. [3 ] I. E. G. Richardson. 2003 H.264 and MPEG-4 Video Compression.
John Wiley Publisher.
[4 ] LI Man H. 2006 Variable Block Size Motion Estimation Hardware for Video Encoders.
[5 ] M. Haller, A. Krutz, and T. Sikora. 2009 Evaluation of pixel- and motion vector-based global motion estimation for camera motion characterization.
[6 ] M.Ritch and N. Canagarajah(2007), “Motion-based video object tracking in the compressed domain,” in Proc. IEEE ICIP,301-304. [7 ] Prabhakar.Telagarapu, V.Jagan Naveen, A.Lakshmi..Prasanthi,
G.Vijaya Santhi.2011 Image Compression Using DCT and Wavelet Transformations.
[8 ] R. V. Babu, K. R. Ramakrishnan, and S. H. Srinivasan. 2004 Video object segmentation: A compressed domain approach,” IEEE Trans. Circuits Syst.
[9 ] S. B. Gokturk and A. M. Aaron.2002 Applying 3d techniques to video for compression”, in Digital Video Processing (EE392J) Projects Winter Quarter.