Wyniki wyszukiwania - BazTech

1

Wizyjny algorytm wyznaczania obszarów zainteresowania zrealizowany w oparciu o zredukowaną ilość przechowywanej informacji

Lech P.

Logistyka

|

2014

|

nr 3

3674--3680

PL

W pracy przedstawiono metodę bazującą na technikach wizyjnych wyznaczania obszaru zainteresowania (ang. region of interest – ROI) dla obrazów cyfrowych lub sekwencji wideo. Opisany algorytm bazuje na uproszonej strukturze opisującej pierwotny obraz cyfrowy uzyskany przy pomocy metody Monte Carlo. Obraz podzielony jest na bloki. Opracowana struktura prócz uproszczonej informacji o obrazie może zawierać wstępnie wyznaczone parametry opisujące obraz np. histogram. Poprzez analizę zapisów z poszczególnych klatek wideo zapisanych w takiej strukturze można określić np. obszary o określonej częstości zmian itp. Poprzez odpowiednie kształtowanie kryteriów, na podstawie których dokonujemy wyboru ROI możemy użyć opracowany algorytm do śledzenia ruchomych obiektów (samochodów itp.) bez ich identyfikacji, wyznaczania obszarów do filtracji cyfrowej pierwotnego obrazu lub wyznaczenia potencjalnego miejsca w obrazie, w którym np. znajduje się tablica rejestracyjna samochodu. Opracowany algorytm może współpracować z sekwencjami wideo lub pojedynczym zdjęciem. W związku z działaniem w oparciu o uproszczoną strukturę jest to algorytm o niskim koszcie obliczeniowym umożliwiającym pracę w czasie rzeczywistym wraz z zrównolegleniem obliczeń.

EN

The paper presents a method based on machine vision techniques for determining the region of interest ( ROI ) for digital images or video sequences. The algorithm is based on the simple data structure incoming from the image obtained using the Monte Carlo method. The image is divided into blocks. The simplified structure is prepared for each one. This structure of an image can be used to calculate parameters describing the image such as a histogram. By analyzing the records of each video frame stored in such a structure can be determined, for example areas with some changes. By shaping the criteria by which we choose the ROI we use this algorithm: to track moving objects (cars, etc.) without their identification, designation of regions for digital filtering of the original image, for searching a potential place in the image where is a license plate of the car. The proposed algorithm can work with video sequences or single images. It is an algorithm with low computational cost and permitting to operation in real time. This algorithm is useful for parallel computation.

2

Motion analysis for image sequence coding

Marchewka A.

Image Processing & Communications

|

2006

|

Vol. 11, no 1

101-105

EN

In this article we present with small modification compression standards H.26x new approach to the indexing most important regions in video sequences. This method uses block-based motion vector estimation and random Markov field. In one of section we present a discussion about an influence of block shape on results of segmentation and motion estimation. This work is not completed and many of particular results will be shown later.

3

Object based segmentation of video using variational level sets

Cheng B., Wang Y., Zheng N., Bian Z.

Machine Graphics and Vision

|

2005

|

Vol. 14, No. 2

145-157

EN

The paper demonstrates a new approach to video segmentation which retains some of the attractive features of existing methods and overcomes some of their limitations. The video sequence is represented as a spatio-temporal volume, and is segmented by an extension of active contour model based on Mumford-Shah techniques. The energy function minimization is similar to 3D interface evolution with curvature-dependent speeds. The spatio-temporal volume need not to be smoothed before processing because our method is not sensitive to noise. Each object needs a closed interface, which is embedded as a level set of a higher-dimensional functions, and is propagated by solving a partial differential equation. The interface stops in the vicinity of object boundaries, which are not necessarily defined by the gradient and can be represented with complex topologies. Finally, an experiment is given to show the effectiveness and robustness of the method.

4

Software and methods for steganography

Bailey K., Condeli J., Curran K.

Image Processing & Communications

|

2005

|

Vol. 10, no 2

103-130

EN

The object of steganography is to send a message through some innocuous carrier to a receiver while preventing anyone else from knowing that a message is being sent. The carrier can be one of several forms of digital media however the most common type of carrier used is an image. The image should not attract any attention as a carrier of a message and should look as close as possible to an ordinary image. This paper summarises recent research in the area of steganography with the aim of evaluating the trends in this area and developing a perspective on it. It looks at the application of steganography to digital images. Different ways of hiding the message sender arc discussed. An extensive study of existing steganography software is provided. Watermarking and cryptography are briefly described. Watermarking is the process of hiding information in a carrier in order to protect the ownership of text, music, films and art. Steganography techniques can be used for the purposes of watermarking. It is important also to briefly mention cryptography as it is commonly used along with steganography to generate an encrypted message to be hidden in a carrier.

5

Image motion estimation: a survey

Marchewka A.

Image Processing & Communications

|

2005

|

Vol. 10, no 1

5-12

EN

This article has the review character and it is an introduction to research in motion analysis used to video sequence compression. In this publication all algorithms are classified into three groups: change detection in scene, characteristic point in image and optical flow. From the group mentioned above, the method that fits best for use in telecommunication systems has emerged.