Precise Eye Gaze Tracking in Video Frames via Combining Particle Filters Method with Genetic Algorithm
سمیرا
افصحی
دانشگاه آزاد اسلامی واحد اسلامشهر - آموزشکده فنی و حرفه ای سما
author
مجتبی
جعفری
دانشگاه آزاد اسلامی واحد اسلامشهر - آموزشکده فنی و حرفه ای سما
author
text
article
2017
per
Gaze tracking is a key process in human-computer interactions. A basic approach in this field is to predict the position of the pupil in sequential video frames. Particle filter, the most important method which is based on aforementioned approach, has a low precision in predicting the pupil position though guaranteeing expected speed. To solve this problem, in this paper, genetic algorithm (GA) is used in the sampling step of the particle filter method. As a result, in each frame, diversity of the particles required for predicting the pupil position in the next video frame is preserved and at the same time the monotony of them is reduced. To evaluate performance, corresponding to different particle populations, speed and precision of the proposed method and the basic particle filter method in predicting pupil positions in video frames of eye were computed and compared. Results show that the proposed method, compared to the basic particle filter method, tracks the gaze more precisely in a lower time.
Advanced Signal Processing
Vice Chancellery for Research and Technology, University of Tabriz
2676-3397
1
v.
1
no.
2017
1
14
https://jasp.tabrizu.ac.ir/article_5527_a6822fd9782d52a6c6d867b67cc73f37.pdf
dx.doi.org/10.22034/jasp.2017.5527
Using Genetic and Hill Climbing Algorithms to Improve Performance of Image Steganography
Vahideh
Rashidzadeh
Faculty of Electrical and Computer Engineering, University of Tabriz
author
Mohammad Ali
Balafar
Faculty of Electrical and Computer Engineering, University of Tabriz
author
Seyed Naser
Razavi
Faculty of Electrical and Computer Engineering, University of Tabriz
author
text
article
2017
per
In this paper, a new method based on genetic and hill Climbing algorithms for steganography in image is proposed. Data is embedded in LSB. In order to enhance image quality and storage capacity, genetic algorithm is utilized which model hiding as a search and optimization process. Also, to speed up the search function with genetic algorithm, local search and hill Climbing algorithms have been used. The findings show that the hiding image quality is acceptable while capacity is increased. Also, hill Climbing Algorithm speed up the hiding process.
Advanced Signal Processing
Vice Chancellery for Research and Technology, University of Tabriz
2676-3397
1
v.
1
no.
2017
15
25
https://jasp.tabrizu.ac.ir/article_5528_b228f35589d87b3f7bc6bf94e4df830b.pdf
dx.doi.org/10.22034/jasp.2017.5528
On the Effects of Flat Fading Channels on the Steady-State Performance of Diffusion Adaptive Networks
Azam
Khalili
Department of Electrical Engineering, University of Malayer
author
Amir
Rastegarnia
Department of Electrical Engineering, University of Malayer
author
Vahid
Vahidpour
Department of Electrical Engineering, University of Malayer
author
Tohid
Yousefi Rezaii
Faculty of Electrical and Computer Engineering, University of Tabriz
author
text
article
2017
per
Adaptive networks are known as powerful solution for distributed estimation problems. It is shown in available works that, under the ideal link condition, diffusion adaptive networks are efficient solutions for distributed estimation. However, ideal link is not a practical assumption for many applications. Thus, this paper aims to study the steady-state performance of diffusion adaptive networks with flat fading channels. Using the energy conservation argument, we derive closed-form expressions for EMSE and MSD metrics. We also derive the required bound (in terms of the step size parameter) for stability of diffusion adaptive network with fading links. Our analysis shows that in this condition, steady-state curves are not monotonic increasing functions of step size. We provide simulation results to support the analysis.
Advanced Signal Processing
Vice Chancellery for Research and Technology, University of Tabriz
2676-3397
1
v.
1
no.
2017
27
34
https://jasp.tabrizu.ac.ir/article_5530_2df99be9d138cd3d5e4e9dc990a31754.pdf
dx.doi.org/10.22034/jasp.2017.5530
Semi-Fragile Watermarking for Image Tamper Detection and Self-Recovery Based on Discrete Cosine Transform
Behrooz
Boloorian Haghighi
Computer Engineering, Faculty of Engineering, Ferdowsi University of Mashhad
author
Amir Hossein
Taherinia
Computer Engineering, Faculty of Engineering, Ferdowsi University of Mashhad
author
text
article
2017
per
Digital watermarking technique embeds additional data (watermark) into digital contents in such a way that the resulting distortion is imperceptible. Tampered detection and recovery is one of the digital watermarking applications. Tamper detection and recovery can mine the tampered region in the image and then recover it. This application can resolve the problem of digital content modification. In this paper, a novel watermarking approach is proposed for tamper detection and recovery in the compressed and uncompressed images. In this method, each block in the image contains watermark of another block. Therefore, for each non-overlapping block in the image, there exists one copy of the watermark. Here, the Chaotic map is used to improve security and confuse block. Also, a secret key is transmitted along with the watermarked image. In our algorithm, a 90% tampered image can be recovered by PSNR ≈ 20 dB. Experimental results demonstrate that the proposed method is secure and superior to the other techniques, especially when the tampered area is large and is resistant to compression.
Advanced Signal Processing
Vice Chancellery for Research and Technology, University of Tabriz
2676-3397
1
v.
1
no.
2017
35
46
https://jasp.tabrizu.ac.ir/article_5531_f25856d598bb7cfe11eebfb5a6572864.pdf
dx.doi.org/10.22034/jasp.2017.5531
Analysis of Echo Signal in Tracking Radar by Time Histogram and Its Experimental Testing
Mohsen
Mivehchy
Faculty of Engineering, Department of Electrical Engineering University of Isfahan, Isfahan
author
Roza
Banitalebi Dehkordi
Faculty of Engineering, Department of Electrical Engineering University of Isfahan
author
Abolghasem
Zeidaabadi Nezhad
Department of Electrical & Computer Engineering Isfahan University of Technology
author
text
article
2017
per
Generally the fractional bandwidth of tracking radar echo signal is lower than an octave. The existence of the target echo signal creates a peak in the histogram of periods at the frequency of the signal. When there is only white noise in the received signal, no distinct peak can be observed in the histogram. However, when both weak echo signal and white noise are coexisted, a peak appears which grows as signal to noise ratio (SNR) increases. This paper presents a simple way based on direct analysis of periods in time domain to distinguish echo signal with low SNR. Comparison of the results from simulation with those from measurement confirms the accuracy of the suggested method.
Advanced Signal Processing
Vice Chancellery for Research and Technology, University of Tabriz
2676-3397
1
v.
1
no.
2017
47
53
https://jasp.tabrizu.ac.ir/article_5532_d095a68598754244ed75dcfed9c16e8f.pdf
dx.doi.org/10.22034/jasp.2017.5532
A Combined Method for Content Based Image Resizing
Farzin
Yaghmaee
Electrical and Computer Engineering Department, Semnan University
author
Kimia
Peyvandy
Electrical and Computer Engineering Department, Semnan University
author
text
article
2017
per
Nowadays there are many different devices for displaying images. According to using images in different applications and various devices, displaying the images in different display size is vital and essential. Therefore it is very important to preserve content of images during resizing. In this paper, we present two different methods for content aware image resizing that combined with Seam carving algorithm and scaling. In both methods, at first the important regions of image are determined. After that Seam carving algorithm is used in these important regions and traditional scaling is applied for other regions. The proposed approaches are implemented and the experimental result shows that these methods are appropriate for various images especially in content images with many objects and edges.
Advanced Signal Processing
Vice Chancellery for Research and Technology, University of Tabriz
2676-3397
1
v.
1
no.
2017
55
64
https://jasp.tabrizu.ac.ir/article_5533_348709219f18c0a632d03f69e7605470.pdf
dx.doi.org/10.22034/jasp.2017.5533