دانلود مقالات ششمین کنفرانس ماشین بینایی و پردازش تصویر ایران - صفحه 3 - Artificial Intelligence

**Reyhane** · ۰۸-۱۷-۱۳۹۰, ۰۱:۱۴ بعد از ظهر

78- ماشين بردار پشتيبان چند شاخه اي واعمال آن در طبقه بندي و آشكارسازي پوست

خلاصه مقاله:
اين مقاله به ارائه يك درخت تصميم مبتني بر تابع هزينه ماشين بردار پشتيبان مي پردازد استفاده از حاشيه اطمينان در طبقه بندي روش پيشنهادي را نسبت به درخت تصميم c5 كاراتر كرده است از طبقه بند پيشنهادي كه نام ماشين بردار پشتيبان چند شاخه اي بر ان نهاده ايم در شناسايي و استخراج قانون از روي داده ها پرداخته مي شود چند مجموعه داده از بانك اطلاعاتي uci انتخاب مي گردد. مقايسه روش پيشنهادي با روش c5 برتري آن را نشان ميدهد ارائه يك مفهوم عدم طبقه بندي در روش پيشنهادي براي نمونه هاي ازمون اين امكان را ميدهد تا از طبقه بندهاي ديگر به شكل سلسله مراتبي در جهت افزايش كارايي سيستم بهره برد كه به كار اتي سپرده مي شود هزينه كم در ايجاد درخت با تابع هزينه ماشين بردار پشتيبان نسبت به c5 كه از انتروپي و جستجوي كلي و خوشه بندي بهره مي برد از ديگر برتري هاي روش ارائه شده است استفاده از روش پيشنهادي در استخراج قوانين در شناسايي پوست نتايج جالبي روي تصاوير ميدهد كه حاكي از قوانين ساده قابل فهم براي بشر و كم روي تصاوير رنگي به دست ميدهد سادگي بكارگيري روش ارائه شده در آشكارسازي هدف درصحنه هاي ويدئويي نويد استفاده آن توسط ديگر محققين است.

كلمات كليدي:
آشكارسازي پوست انتروپي، درخت تصميم c5، ماشين بردار پشتيبان، طبقه بندي

**Reyhane** · ۱۰-۳-۱۳۹۰, ۰۴:۱۶ بعد از ظهر

140 - A Novel Approach for Fast and Robust Multiple License Plate Detection

License Plate Detection (LPD) is the most difficult,critical and time consuming task in license plate recognition (LPR) systems. In this paper, a novel texture-based method is proposed for fast and robust LPD. First, a new filter called Peak-Valley filter is applied on the lines of the image. This filter not only extracts the remarkable gray level changes as consecutive peaks and valleys, but also simultaneously removes the undesirable small variations. Secondly, a sequential Peak-Valley partitioning is utilized to segment the transitions into some groups. Afterward, a neural network is employed to find true candidate lines and finally the candidate lines are aggregated to form the plates regions. According to our experiments, the proposed method correctly detects all plates presented in the image regardless of their styles and without considering the whole image. Experimental results showed that this approach can apply on real-time application for outdoor complex scenes.

**Reyhane** · ۱۰-۳-۱۳۹۰, ۰۴:۲۷ بعد از ظهر

162- Computer-aided Detection of Proliferative Cells and Mitosis Index in Immunohistichemically Images of Meningioma

Immuonohistochemically images of meningioma which are stained by ki66 marker contain positive and negative cells. Accurate counting the number of positive and negative cells in such images play a critical role in diagnosing diffrent type of meningioma cancer. Since pathological images of meningioma contain complex cell cluster accurate cell counting methodology is a major challenge for pathologist physicians. In this paper we provide a computer aided algorithm for detecting proliferative cells and mitosis index in immunohistochemically images of meningioma. In the first stage of the algorithm fuzzy c-means clustering was used to extract positive and negative cells based on CIElab color space. In the second stage, ultraerosion operation was applied to count the number of individual and overlapped cells. Experimental result show that the proposed algorithm is able to overcome some disadvantage of traditional approaches with acceptable accuracy by pathologist physicians.

**Reyhane** · ۱۰-۳-۱۳۹۰, ۰۵:۴۰ بعد از ظهر

164- Detection and Classification of Foreign Substances in Medical Vials Using MLP Neural Network and SVM

خلاصه مقاله:
Presence of foreign substances in medical liquids can make serious problems for both patients and companies. To avoid these problems, there is a vast need of an automatic process to identify the bottles with foreign substances. In this paper, a new method is proposed to detect and classify the foreign substances in medicine bottles and vials based on machine vision. Several cameras are located in production line, to get images from medicine bottles. The captured images are thresholded to gather a collection of connected components. For each one a set of novel features are computed, the feature vectors are fed into a classifier, to distinguish the foreign substances from bubbles and also classify them in four groups, so the operator can find the source of the problem and fixes the failure in machine which causes it.. An original method is also described to find out the scratches and spots on the bottle surface and distinguish them from foreign substances. The proposed method achieves detection rates over 97% and classification rates over 93%.

**Reyhane** · ۱۰-۳-۱۳۹۰, ۰۵:۴۹ بعد از ظهر

189- Real time and robust intersection traffic flow measurement using traffic zones in a hybrid background modelling for smart cameras

خلاصه مقاله:
Background modelling methods for identifying Vehicles in a traffic video surveillance is a fundamental task in computer-vision applications. In intelligent transportation systems (ITS), traffic parameters extraction at intersections is one of the critical and challenging tasks in urban traffic management. For intersection traffic analyzing where objects have different characteristics such as varying velocities, stop and go, it is necessary to use the adaptive background mixture model to learn background model faster and more accurately, instead of using single rate of adaptation, which is not adequate. The main focus of this research is to analyze activities at intersection for detecting and classifying vehicles and then extract traffic flow which assists in regulating traffic lights for using in a smart camera. Traffic zones definition in intersection video based on majority motions, greatly reduce the computations. A smart camera’s fundamental purpose is to analyze a scene and report statics and activities of interest which is not an image. Ground-truth experiments with urban traffic sequences show that our proposed algorithm is very promising relative to results using other techniques.

**Reyhane** · ۱۰-۳-۱۳۹۰, ۰۶:۱۴ بعد از ظهر

190- Real-Time Fusion of Multi-Focus Images for Visual Sensor Networks

خلاصه مقاله:
The objective of image fusion is to combine information from multiple images of the same scene in order to deliver only the useful information. The discrete cosine transform (DCT) based methods of image fusion are more suitable and time-saving in real-time systems using DCT based standards of still image or video. In this paper an efficient approach for fusion of multi-focus images based on variance calculated in DCT domain is presented. The experimental results on several images show the efficiency improvement of our method both in quality and complexity reduction in comparison with several recent proposed techniques

**Reyhane** · ۱۰-۳-۱۳۹۰, ۰۷:۴۹ بعد از ظهر

193- Stationary Image Resolution Enhancement on the Basis of Contourlet and Wavelet Transforms by means of the Artificial Neural Network

خلاصه مقاله:
In this paper two transform based super resolution methods are presented for enhancing the resolution of a stationary image. In the first method, neural network is trained by wavelet transform coefficients of lower resolution of a given image, and then this neural network are used to estimate wavelet details subbands of that given image. In this way, by using these estimated subbands as wavelet details and the given image as the approximation image, a super-resolution image is made using the inverse wavelet transform. In the second proposed method, the wavelet transform is replaced by contourlet transform and the same mentioned procedure is applied. These two methods have been compared with each other and with the bicubic method on different types of images. The experimental results demonstrate the superiority performance of the proposed methods compared with regular stationary image resolution enhancing methods.

**Reyhane** · ۱۰-۳-۱۳۹۰, ۰۷:۵۲ بعد از ظهر

200- Vector-Valued Color Image Edge Detection Using Green Function Approach

خلاصه مقاله:
In this paper, an extended version of differential equations is introduced on the concept of edge detection methods for color images based on the correlation of R, G, and B components. To obtain the color edge detection operator, the Green’s function approach is used to derive the obtained differential equation. The proposed color edge detection method is compared with other color edge detection methods on the several test images. The experimental results show the feasibility of the proposed approach.

**Reyhane** · ۱۰-۳-۱۳۹۰, ۰۷:۵۶ بعد از ظهر

177- Improved Motion Compensated Temporal Filtering

خلاصه مقاله:
The new MPEG standardization effort of H.462 /AVC scalable extension, Joint Scalable Video Coding (JSVM) is currently being made for efficient scalable video coding (SVC). Reducing the energy of high-pass temporal frames helps to improve the quality and supports for compression efficiency. Since bi directional motion–compensated prediction generally reduces the energy of the prediction residual in comparsion to uni-directional prediction, thus we use the bi directional prediction and by selecting the coefficients of prediction step, adaptively, we are going to reduce the energy of high-pass temporal frames. Experimental results show that the proposed algorithm can significantly improve the quality of the reconstructed video sequence, both in PSNR and in subjective visual quality. as proposed technique can improve coding performance by up 3.0 dB.

**Reyhane** · ۱۲-۶-۱۳۹۰, ۰۵:۰۰ بعد از ظهر

99- تشخيص و بازشناسي الفباي دستي گوياي باغچه بان و اعداد زبان اشاره فارسي و استفاده از جداسازي مبتني بر قاعده تكنيكهاي بينايي ماشين و شبكه عصبي

خلاصه مقاله:
انسانهادر زندگي روزانه براي برقراري ارتباط راحت تر با يكديگر از فرمانهاي اشاره اي دست به كرات استفاده مي كنند يكي از وسايل موثر درمحاوره هاي انساني ارتباط به كمك ايما و اشاره مي باشد كه بهترين نوع آن اشاره به كمك دستان است زبان اشاره به وسيله ارتباطي براي افرادي است كه قادر به شنيدن نيستند و يا در محيطي قرار دارند كه به دلايل خاص محيطي مجبور به استفاده از اين زبان هستند اما بيشترين استفاده اين حركات براي افرادي است كه از نعمت شنيدن بي نصيب مانده اند اين زبان مجموعه اي از ژست ها و حركات است كه مي توان با آنها حروف، كلمات، و يا اعداد را بيان كرد اينمقاله نحوه تشخيص و بازشناسي الفباس دستي گوياي باغچه بان و اعداد را ارائه مي كند بدين صورت كه با استفاده از تكنيكهاي بينايي ماشين دست ها و سر يك انسان رادر تصوير مشخص و استخراج مي كنيم سپس با استفاده از قواعدي اين تصاوير را دسته بندي مي كنيم و با استفاده از ويژگيهاي استخراج شده از آنها يك شبكه عصبي رااموزش ميدهيم و از اين شبكه براي ازمايش حركات اشاره استفاده مي كنيم.

كلمات كليدي:
الفباي دستي گويايي باغچه بان، جداسازي مبتني بر قاعده، شبكه عصبي ، فضاي رنگ، عمليات شكل شناسي