Model-Based JPEG2000 rate control methods

Aquesta recerca esta centrada en l'escalabilitat qualitativa de l'estandard de compressio d'imatges JPEG2000. L'escalabilitat qualitativa es una caracteristica fonamental que permet el truncament de la tira de bits a diferents punts sense penalitzar la qualitat de la imatge recuperada. L'escalabilitat qualitativa es tambe fonamental en transmissions d'imatges interactives, ja que permet la transmissio de finestres d'interes a diferents qualitats. El JPEG2000 aconsegueix escalabilitat qualitativa a partir del metode de control de factor de compressio utilitzat en el proces de compressio, que empotra capes de qualitat a la tira de bits. En alguns escenaris, aquesta arquitectura pot causar dos problematiques: per una banda, quan el proces de codificacio acaba, el numero i distribucio de les capes de qualitat es permanent, causant una manca d'escalabilitat qualitativa a tires de bits amb una o poques capes de qualitat. Per altra banda, el metode de control de factor de compressio construeix capes de qualitat considerant la optimitzacio de la rao distorsio per l'area completa de la imatge, i aixo pot provocar que la distribucio de les capes de qualitat per la transmissio de finestres d'interes no sigui adequada. Aquesta tesis introdueix tres metodes de control de factor de compressio que proveeixen escalabilitat qualitativa per finestres d'interes, o per tota l'area de la imatge, encara que la tira de bits contingui una o poques capes de qualitat. El primer metode esta basat en una simple estrategia d'entrellacat (CPI) que modela la rao distorsio a partir d'una aproximacio classica. Un analisis acurat del CPI motiva el segon metode, basat en un ordre d'escaneig invers i una concatenacio de passades de codificacio (ROC). El tercer metode es beneficia dels models de rao distorsio del CPI i ROC, desenvolupant una novedosa aproximacio basada en la caracteritzacio de la rao distorsio dels blocs de codificacio dins una subbanda (CoRD). Els resultats experimentals suggereixen que tant el CPI com el ROC son capacos de proporcionar escalabilitat qualitativa a tires de bits, encara que continguin una o poques capes de qualitat, aconseguint un rendiment de codificacio practicament equivalent a l'obtingut amb l'us de capes de qualitat. Tot i aixo, els resultats del CPI no estan ben balancejats per les diferents raons de compressio i el ROC presenta irregularitats segons el corpus d'imatges. CoRD millora els resultats de CPI i ROC i aconsegueix un rendiment ben balancejat. A mes, CoRD obte un rendiment de compressio una mica millor que l'aconseguit amb l'us de capes de qualitat. La complexitat computacional del CPI, ROC i CoRD es, a la practica, negligible, fent-los adequats per el seu us en transmissions interactives d'imatges. This work is focused on the quality scalability of the JPEG2000 image compression standard. Quality scalability is an important feature that allows the truncation of the code-stream at different bit-rates without penalizing the coding performance. Quality scalability is also fundamental in interactive image transmissions to allow the delivery of Windows of Interest (WOI) at increasing qualities. JPEG2000 achieves quality scalability through the rate control method used in the encoding process, which embeds quality layers to the code-stream. In some scenarios, this architecture might raise two drawbacks: on the one hand, when the coding process finishes, the number and bit-rates of quality layers are fixed, causing a lack of quality scalability to code-streams encoded with a single or few quality layers. On the other hand, the rate control method constructs quality layers considering the rate¬distortion optimization of the complete image, and this might not allocate the quality layers adequately for the delivery of a WOI at increasing qualities. This thesis introduces three rate control methods that supply quality scalability for WOIs, or for the complete image, even if the code-stream contains a single or few quality layers. The first method is based on a simple Coding Passes Interleaving (CPI) that models the rate-distortion through a classical approach. An accurate analysis of CPI motivates the second rate control method, which introduces simple modifications to CPI based on a Reverse subband scanning Order and coding passes Concatenation (ROC). The third method benefits from the rate-distortion models of CPI and ROC, developing an approach based on a novel Characterization of the Rate-Distortion slope (CoRD) that estimates the rate-distortion of the code¬blocks within a subband. Experimental results suggest that CPI and ROC are able to supply quality scalability to code-streams, even if they contain a single or few quality layers, achieving a coding performance almost equivalent to the one obtained with the use of quality layers. However, the results of CPI are unbalanced among bit-rates, and ROC presents an irregular coding performance for some corpus of images. CoRD outperforms CPI and ROC achieving well-balanced and regular results and, in addition, it obtains a slightly better coding performance than the one achieved with the use of quality layers. The computational complexity of CPI, ROC and CoRD is negligible in practice, making them suitable to control interactive image transmissions.