WELCOME MESSAGE. I wish you all have a successful conference. Putu Angga Aditya Yoga ISCECE 2017 Chair

Transcription

1

2 WELCOME MESSAGE My pleasure to welcome all distinguished participants the 2017 International Student Conference on Electrical and Computer Engineering (ISCECE 2017) which is held on 25 th August 2017, in Wiswacarma Hall, Faculty of Engineering, Udayana University. The ISCECE 2017 is organized by IEEE Udayana University Student Branch. The conference is conducted in conjunction with the 55 th Anniversary of Udayana University and the 52 nd Anniversary of Faculty of Engineering The theme of this conference is Building Smart Generation Toward Green Society. Building smart young generation to create innovations in order to become green society. The conference has received 10 submissions. There are 5 papers for presentation but one papers was rejected, so there are 9 papers in this proceeding. Thank you to our keynote speaker Prof. Kouichi Takase of Nihon University and our host speaker Yoga Divayana, Ph.D of Udayana University for their support in this conference. I would like to thank Department of Electrical Engineering, Faculty of Engineering, Udayana University for their support and funding. I am gratefully for the all committees of the International Student Conference on Electrical and Computer Engineering 2017 for their support to prepare this conference. Finally many thanks for all those participants. Without all of your supports, the conference couldn t be held. I wish you all have a successful conference Putu Angga Aditya Yoga ISCECE 2017 Chair i

3 ORGANIZING COMMITTEE Chair : Putu Angga Aditya Yoga Co-Chair : Ni Putu Lintang Anggitiadewi Secretary : Ida Bagus Vidananda Agastya Made Candra Adi Sutra Nia Paramitha Finance : Ida Ayu Sri Sinta Anjani Program : Putu Ratih Devyanti I Wayan Narayana Putra Mochammad Farhan Arieffadillah Desak Komang Kharisma Jandinhi Komang Agus Angga Prasetiawan Michael Candra Santoso Wayan Gede Yoga A.W Joshua Fernaldy Sudarsono Made Dwi Wiprayoga Secretariat : Putu Diana Sari I Ketut Ary Suarjaya Putra I Wayan Krisna Saputra ii

4 Ni Komang Ayu Sri Anggreni Gede Teguh Pradnyana Yoga I Putu Weda Jayanthana Sponsorship : Ni Made Anita Belinda Ni Made Neli Lestari Agus Satrya Wibawa I Gede Wiyoga Putra I Putu Eka April Yanto I Nyoman Adi Guna Subawa Anak Agung Gede Agung Semarabawa I Gede Hery Putrawan Publication And Documentation : Komang Triadi Antara Siti Ayu Oktavianti Putra Yudhanata Pratama I Gede Bayu Suarsa Yoga Kusuma Wardhana iii

5 TABLE OF CONTENTS Welcome Message... i Organizing Committee... ii Table of Contents... iv Analyze the Effect of Delay Spread on MIMO Zero Forcing System on Selective Fading Channel... 1 IG Darma Putra, IGAK Diafari Djuni, NMAED Wirastuti Field Study at Seloliman Village in East Java using Kali Maron Micro Hydro Power Plants as Power Supply... 8 Haksari Laksmi Bestari, I N Satya Kumara Overview of Integrated Monopole Microcell in Bali Rahmat Rudiantono, I Made Indra Wiguna, Linawati Performance Analysis of Site G_3G_KUTRI On Football Matches at Dipta Gianyar Stadium I Made Indra Wiguna, Widyadi Setiawan Development Of Piezoelectric (Piezoelectric Utilization On The Street To Reducing Petroleum Usage) Kadek Bintang Anjasmara, I Nyoman Sumitra Tanaya, Nyoman Budiastra, The Calculation of PSNR Value on Extended Bilateral Motion Estimation and Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation Muhammad Iman Nur Hakim, N. M. A. E. D. Wirastuti IoT Based Swimming Pool Stabilizer Using Wemos D1 Mini Luh Ayu Sutawati, Made Surya Kumara, Nyoman Wawan Sandi Prayoga, Cok Gede Indra Partha iv

6 Alternative Energy Sources Using Thermoelectric Modules I Made Surya Kumara, Nyoman Wawan Sandi Prayoga, Luh Ayu Sutawati, Gede Sukadarmika Designing and Building a Device to Prevent Sexual Crime Nyoman Wawan Sandi Prayoga, Ni Kadek Danis Lisyaningsih, Luh Ayu Sutawati, Marcellino Adi Chrisna Moeri, Ida Bagus Putu Teguh Brahmantika, I Gusti Ngurah Janardana Author Index v

7 The Calculation of PSNR Value on Extended Bilateral Motion Estimation and Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation Muhammad Iman Nur Hakim 1, N. M. A. E. D. Wirastuti 2 1,2 Postgraduate of Electrical Engineering, Udayana University Denpasar, Indonesia m.iman.nh@gmail.com dewi.wirastuti@ee.unud.ac.id Abstract The requirement for multimedia information increase, but the limitations of the network becomes a problem in the development of multimedia. To overcome this, we need an image or video processing techniques so that image and video sent does not require a big sources. One of the techniques is Extended Bilateral Motion Estimation and Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation that will be tested on Akiyo video and Container video. The results of this research are average value of PSNR on Akiyo video with Extended Bilateral Motion Estimation methods is greater than the results of Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation method. But in a Container video, Extended Bilateral Motion Estimation methods produce an average value lower than Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation methods although indeed differences in value is not too big either on Akiyo video and Container video. Keywords PSNR; Extended Bilateral Motion Estimation; Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation. I. INTRODUCTION With the growing information technology makes images and video delivery request, had to be considered. Moreover, that is currently user are not just looking for information in text form, but rather in the form of pictures and video. However, there are obstacles that must be faced for sending pictures and videos require network resources fairly high, while the channels is quite limited. With the limitations of the network, it is necessary to do an image processing techniques for the image or video so it can still be sent. However, although has undergone image processing or video processing, image and videos still contain the original information therein and not much different from the original source. The techniques about image and video processing is a Extended Bilateral Motion Estimation and Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation. This technique is used to reduce the required capacity during transmission and at the decoder, the video will be scaled back to resemble the quality of information sources. Extended Bilateral Motion Estimation is one where motion estimation compression technique will make the process for determining the movement of an object on a digital video sequences. The movement is generally represented in the form of motion vector at selected points in the current frame compared to other frame is referred as a reference frame [1]. While the concept of wavelet image compression so even though the image data is already compressed (compression) does not eliminate the existing information. Besides being used in a stationary image compression, wavelet is also widely used in several other studies, such as the moving image compression, remote sensing, and detection of computer network activity [2]. II. IMAGE AND VIDEO PROCESSING A. Extended Bilateral Motion Estimation Motion Estimation is a process to determine the movement of objects in video sequences. Motion Estimation is one Interframe compression techniques that predict a frame from the previous frame (reference frame), by estimating the movement of blocks between frames. The frame is divided into blocks that do not overlap. Each block is compared to equally sized blocks, on previous frame by block matching In conducting these match, the location of the most similar block or match on reference frame different from the location of the target block. Differences in the relative positions is called the motion vector (MV), as shown in Fig. 1. If the position of the target block and the match block are the same, then the motion vector is zero. This motion vector that indicates a shift in the blocks between frames. When encodes each block of a frame that is predictable, vector motion which indicates the position of the block match on reference frame, encoded in the position of the target block itself, then there is compression, because the number of bits needed to encode the vector of motion less than encode a block as a whole. To improve the accuracy of conventional BME, the extended BME (EBME) [3] extra calculates MV of the block overlapping with each of two adjacent original blocks by half to find the true MV with higher probability than the conventional BME, the dual ME [4] uses the unidirectional and bidirectional matching ratios of blocks in the previous and 27

8 following frames to refine the motion vector field (MVF) of interpolated frame, and the direction-select ME (DS-ME) [5] independently calculates the forward and backward MVs of the interpolated frame and then selects the more reliable one from them by using the sum of bilateral absolute differences (SBAD). Fig. 1. Motion Vector B. Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation. Wavelet transformation is a method for selecting data, function or operator into components of different frequencies, then identify each component at a resolution that matches the scale. The flowchart of Wavelet Pyramid Based Multi- Resolution Bilateral Motion Estimation illustrated in Fig. 2. With each of the previous frame f t-1 and following frame f t+1. Each level of wavelet pyramid consists of two parts: sub sampled image and its edge image, and therefore and edgedependent term can be included in the block matching criterion to improve the accuracy of Motion Estimation (ME). After the construction of wavelet pyramids, the full-search based BMA is applied to the top-level images (i.e., the images with the lowest resolution) for obtaining the initial MVF of the interpolated frame. [6]. To maintain the same density of MVF at each level, the blocks size of sub sampled decreases by two times. The illustrated of blocks between two pyramids level, it show in Fig. 3. The bottom-level image (i.e., the input original video frame) is decomposed into four lower resolution components which are respectively called LL sub-band, LH sub-band, HH sub-band and HL sub-band in the clock-wise direction. The LL sub-band contains the low frequency components of the bottom-level image, and it is extracted from wavelet coefficients to produce the sub-sampled image in the next level. The horizontal, vertical and diagonal directionality of the bottom-level image is clearly presented in the LH, HL and HH sub-bands, and therefore they can be combined into a single edge image by simply zeroing the LL sub-band in wavelet coefficients and then performing the inverse wavelet transform. [6]. Fig. 2. Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation Flowchart Although the wavelet pyramid multi-resolution search can result in a smoother MVF, the mismatch still exists, which introduces the problem that the mismatch will spread at a lower level when it occurs at a certain level. To alleviate the undesirable effect from the above problem, there add the motion vector smoothing before motion vector refinement at each level to guarantee the reliability of predictive MVs so as to prevent the spread of MV errors. Fig. 3. Blocks between two pyramid levels. 28

9 III. PSNR VALUES Peak Signal to Noise Ratio (PSNR) is the ratio between maximum value of the signal measured by the amount of noise that affects the signals. PSNR has a value in decibels (). [7] PSNR is used to compare the image quality before and after the image or video processing. In an image reconstruction, there required comparison between images reconstructed with the original image. Generally used in this case is PSNR. The high of PSNR value states similarity between the results of the reconstruction with the original image. If obtained PSNR values below 30, it can be stated that the quality is relatively low, where the distortions due to the insertion clearly visible. However, with the value of 40 or higher, it can be stated that the quality of the output relatively high. A. PSNR Value with EBME. In this research, used Akiyo video and Container video with cif quality. Differences between these videos are on the subject of motion in the video. Container video contains more motion than Akiyo video. From the experimental results, obtained each PSNR in Fig. 4 for EBME with Akiyo video, Fig. 5 for EBME with Container video, Fig. 6 for Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation with Akiyo video, and Fig. 7 for Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation with Container video. In the research using EBME, on Akiyo video obtained the highest PSNR value at and the lowest at PSNR. In Fig. 4 is shown PSNR on each frame. Fig. 5. PSNR with EBME on Container video. B. PSNR Value with Wavelet Pyramid Based Multi- Resolution Bilateral Motion Estimation Using the Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation method on Akiyo video, the highest PSNR value is and the lowest PSNR value is Each PSNR value from every frame can be seen in Fig. 6. It is not too much different between the result in Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation and EBME. Fig. 4. PSNR with EBME on Akiyo video. In Fig. 5 is shown PSNR value each frame of Container video. The maximum PSNR values obtained in this research at and a minimum value at PSNR. The average PSNR with EBME on the both are for Akiyo video and for Container video. Fig. 6. PSNR with Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation on Akiyo video. In the Container video, the highest PSNR value is and the lowest PSNR value is In Fig. 7 shown PSNR value in each frame. From the resulting of graph, looks not much different from previous results on EBME method. This is close with the result of the Akiyo video. In the research of Wavelet Pyramid Based Multi-Resolution Bilateral Motion 29

10 Estimation obtained PSNR average value for Akiyo video at and for Container video. movement of the subject, produced greater by Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation method than EBME method. Although the difference in value is not too significant. Similarly, the highest nor the lowest PSNR, resulting in greater in Wavelet Pyramid Based Multi- Resolution Bilateral Motion Estimation than EBME even though the value is small. Overall, the Akiyo video obtained better results with EBME method, whereas in the Container video obtained better results by using Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation. Fig. 7. PSNR with Wavelet Pyramid Based Multi-Resolution Bilateral Motion Estimation in Container video. The results of PSNR values have been obtained, we can show in the form of a table as shown in Table 1. The highest PSNR value in video Akiyo have similarities in both methods. Neither the method nor the Wavelet Pyramid Based Multi- Resolution Bilateral Motion Estimation. The highest PSNR value at PSNR Highest PSNR Lowest PSNR Average PSNR TABLE I. Akiyo EBME Wavelet Pyramid PSNR VALUES Container EBME Wavelet Pyramid REFERENCES [1] P.E. Priyatna Panja, Penggunaan motion estimation untuk perbaikan gerakan objek pada kompresi video [The use of motion estimation for repair object on the movement of video compression], [2] Tb. Ai, Munandar, M. Adelvin L., Alb. Joko Santoso, Analisa PSNR, rasio kompresi warna dan MSE terhadap kompresi image menggunakan 31 fungsi wavelet. [Analysis of PSNR, color compression ratio and MSE against image compression using 31wavelet function], Atma Jaya University, Yogyakarta. March, [3] S.-J.Kang, K.-R.Cho, and Y.H.Kim, Motion compensated frame rate up-conversion using extended bilateral motion estimation, IEEE Trans. Consum. Electron., vol.53, no.4, pp , On R. Li, H. Liu, J. Chen, and Z. Gan, Wavelet pyramid based multi-resolution bilateral motion estimation for frame rate up-conversion, no. 1, pp , [4] S.-J.Kang, S.Yoo, and Y.H.Kim, Dual motion estimation for frame rate up-conversion, IEEE Trans. Circuits Syst. Video Technol., vol.20, no.12,pp , On R. Li, H. Liu, J. Chen, and Z. Gan, Wavelet pyramid based multi-resolution bilateral motion estimation for frame rate up-conversion, no. 1, pp , [5] D.-G.Yoo, S.-J.Kang, and Y.H.Kim, Direction-select motion estimation for motion-compensated frame rate up-conversion, J. Display Technol., vol.9, no.10, pp , On R. Li, H. Liu, J. Chen, and Z. Gan, Wavelet pyramid based multi-resolution bilateral motion estimation for frame rate up-conversion, no. 1, pp , [6] R. Li, H. Liu, J. Chen, and Z. Gan, Wavelet pyramid based multiresolution bilateral motion estimation for frame rate up-conversion, no. 1, pp , [7] G.M. Male, Wirawan, E. Setijadi, Analisa kualitas citra pada steganografi e-government [Analysis of image quality on e- government steganography], Proceedings of 15 th the National Seminar on Management of Technology, pp. 1 9, IV. CONCLUSION From the result of research the largest PSNR average value of two methods featured in the Akiyo video than Container video. The average PSNR of both methods on Akiyo video is 46.6, while on video Container is However, the average value of Container video, which contains more 30

11 AUTHOR INDEX A Anjasmara, Kadek Bintang B Bestari, Haksari Laksmi... 8 Budiastra, Nyoman Brahmantika, Ida Bagus Putu Teguh D Darma Putra, IG... 1 Diafari Djuni, IGAK... 1 H Hakim, Muhammad Iman Nur J Janardana, I Gusti Ngurah K Kumara, I Made Surya... 31, 35 Kumara, I N Satya... 8 L Linawati Lisyaningsih, Ni Kadek Danis M Moeri, Marcellino Adi Chrisna P Prayoga, Nyoman Wawan Sandi... 31, 35, 38 Partha, Cok Gede Indra R Rudiantono, Rahmat S Setiawan, Widyadi Sukadarmika, Gede Sutawati, Luh Ayu... 31, 35, 38 T Tanaya, I Nyoman Sumitra W Wiguna, I Made Indra , 18 Wirastuti, NMAED... 1 Wirastuti, N.M.A.E.D