Atıf Sayısı
4
Görüntüleme
19
İndirme
4
Sayısal görüntülerden oluşan bir veri tabanından sorgulanan bir görüntünün aynısının veya benzerlerinin getirilmesi süreci görüntü erişimi olarak tanımlanır. Her ne kadar sayısal görüntü piksellerden oluşuyor olsa da sorgulama piksel düzeyinde değil, sayısal görüntüleri temsil eden vektörler düzeyinde yapılmaktadır. Görüntülerin vektörler ile temsil edilmesi özellik çıkarma süreci olarak adlandırılır ve içerik tabanlı görüntü erişiminin (İTGE) en önemli aşamasıdır. Özellik vektörünün temsil kabiliyetinin düşük olması sistemin performansının da düşük olması demektir. Gri ölçekli görüntülerin histogramları en tipik özellik vektörleridir. Diğer taraftan renkli görüntülerde üç ayrı kanal mevcut olduğundan, görüntüyü temsil edebilecek histogram üç boyutlu bir dizi oluşturur ki bu durum sistemin hesap maliyetini oldukça artıracaktır. Bu nedenle araştırmacılar renkli görüntülerdeki renk sayısını azaltma veya renk indirgeme yaklaşımını tercih etmişlerdir. Vektör kuantalama olarak adlandırılan renk indirgeme sürecinde ise her zaman aynı sonucu üretmek mümkün olmamıştır. Bunun nedeni ise bazı algoritmaların başlangıçta rastgele üretilen renk vektörleri ile çözüm aramalarıdır. Linde-Buzo-Gray (LBG), K-ortalamalar ve bulanık c-ortalamalar algoritmaları bu tür çözüm yaklaşımlarına tipik örneklerdir. Bu çalışmada tekrarlı ortalama tabanlı renk indirgeme yaklaşımı kullanılarak yeni bir görüntü erişim metodu önerilmiştir. Önerilen stratejide, öncelikle her bir renk kanalının histogramı üzerinden tekrarlı bir şekilde ortalamalar hesaplanmış ve çok seviyeli eşikler elde edilmiştir. Elde edilen eşikler kullanılarak RGB renk uzayı alt prizmalar şeklinde dilimlenmiştir. Oluşan alt prizmalar içinde kalan pikseller aynı sınıfa atanmış ve ilgili sınıftaki piksellerin ortalamaları kullanılarak renk indirgemesi yapılmıştır. Sınıf indisleri ve ilgili sınıflara tahsis edilen piksel sayıları yardımıyla tek boyutlu histogram elde edilmiştir. Son aşamada ise elde edilen sınıf tabanlı histogram özellik vektörü olarak atanmış ve içerik tabanlı görüntü erişimi gerçekleştirilmiştir. Önerilen algoritma ve LBG algoritması ile sonuçlar alınmış ve karşılaştırmalar yapılmıştır.
The process of bringing the same or similar of a questioned image from a database consisting of numerical images is defined as image access. Although the numerical image consists of pixels, the query is done not at the level of pixels, but at the level of vectors representing the numerical images. The representation of images with vectors is called the characterization process and is the most important stage of content-based image access (ITGE). The low capacity of the characteristic vector means that the performance of the system is also low. The histograms of gray scale images are the most typical characteristics vectors. On the other hand, since there are three separate channels in colored images, the histogram that can represent the image creates a three-dimensional series that will significantly increase the system’s account cost. Therefore, the researchers have preferred a approach to decrease the number of colors in colored images or decrease the color. In the process of color reduction called vector quantification, it was not always possible to produce the same result. This is due to the fact that some algorithms are searching for solutions with randomly produced color vectors. Linde-Buzo-Gray (LBG), K-mediates and foolish c-mediates algorithms are typical examples of such solutions. In this study, a new image access method was proposed using repeated average-based color reduction approach. In the proposed strategy, the average is calculated repeatedly through the histogram of each color channel and multi-level edges are obtained. The obtained edges are scattered in the RGB color space in the form of subprisms. The remaining pixels in the forming subprisms are assigned to the same class and the color reduction is made using the average of the pixels in the relevant class. A single-dimensional histogram was obtained with the help of class indices and the numbers of pixels allocated to the relevant classes. In the final stage, the class-based histogram characteristics was assigned as a vector and content-based image access was made. With the recommended algorithm and LBG algorithm, the results were taken and compared.
Image retrieval is defined as the process of retrieving same or similar of an image queried from a digital image database. Although a digital image is composed of pixels, the query is not performed at the pixel level but it is carried out at level of vectors representing digital images. In other words, it is computationally necessary to represent with vectors both queried image and images in database. The similarity between queried and database images is computed by vector operations. The representation of images by vectors is called feature extraction process and it is the most significant stage of content-based image retrieval (CBIR). Histograms of gray scale images are typical feature vectors. On the other hand, as there are three different channels in color images, the histograms which represent images are three-dimensional arrays, which will increase the computational cost of the system considerably. For this reason, researchers have preferred to use color quantization or reduce the number of colors in color images. The color reduction process is called as vector quantization, but it is not always possible to produce the same result. The reason is that the developed algorithms so far look for solutions with randomly generated color vectors initially. Linde-Buzo-Gray (LBG), K-means and fuzzy c-means algorithms are typical examples of such solution approaches. In this study, a new image retrieval method has been proposed by using the recursive mean-based color reduction approach. In the proposed strategy, firstly, averages were calculated from the histogram of each color channel and consequently multi-level thresholds were obtained. Using the thresholds obtained, RGB color space was sliced into sub-prisms. The pixels in the created sub-prisms were assigned to the same class and color reduction was made by using the means of pixels in the related class. One-dimensional histogram was obtained with the help of class indices and the number of pixels allocated to the related classes. In the last stage, the obtained class-based histogram was assigned as feature vector and content-based image retrieval was performed. The results were obtained with the proposed algorithm and LBG algorithm. Additionally, comparisons were made.
Alan : Fen Bilimleri ve Matematik
Dergi Türü : Ulusal
Makale Detay 
Atıf Yapanlar
Benzer Makaleler
PDF Görüntüle
Dergi Bilgisi
Eseri Dinleyin
Bu Sayfayı Yazdırın
Paylaş
Mendeley
