Mixed Raster Content for Compound Image CompressionFinal Project Presentation – EE-5359Spring 2009Submitted to:Dr. K.R. RaoSubmitted by:Pritesh Shah(1000555858)MOTIVATION In today’s world it is impossible to imagine a day without information orinformation exchange in digital form over internet. With new advances in data processing systems and scanning devices,documents are present in a wide variety of printing systems. Documents in digital form are easy to store and edit, and can also betransmitted within seconds. These documents may contain text, graphics andpictures. So their storage requires huge memory and also transmission requires highcompression and bit rates to avoid expenses and delay.INTRODUCTION Compound images are documents containing both, binary text and continuoustone images. JPEG can be used for documents containing only pictures and graphics. But whencompressing a compound document, MRC is found to have an upper hand. MRC uses a multi-layer, multi-resolution representation of a compound document. Instead of using a single algorithm, it uses multiple compression algorithms,including the ones specifically developed for text and images. So it can combinethe best of new or existing algorithms and can provide various quality-compression ratio tradeoffs.Mixed Raster Content(MRC) Imaging ModelFig. 1 MRC 3-plane configuration. Foreground(FG) plane poured into Background(BG) through Mask layer[1] FGMASKBG The 3-layer MRC model contains 2 colored image layers(foreground(FG) andbackground(BG)) and one binary image layer(mask)[1]. Mask layer is the decisive layer in reconstructing the image from the FG andBG layers. When the pixel value in the mask layer is 1, the corresponding pixel fromthe FG is selected and when its 0, the corresponding pixel from the BG isselected and the final image is reconstructed as shown in Fig. 1.MRC Framework for Scanned Data[4]Fig. 2 MRC framework In MRC after the original single resolution image is decomposed into layers, theyare processed and compressed using different algorithms. Fig. 2 shows how the original document X is input to the pre-processor whichyields output Y (3-layer MRC model). It also constructs an edge sharpening map and estimates the original edge“softness”. This information is termed as side information. This data is then encoded and then the encoded data goes to the decoder, wherethe reconstructed data Y’, along with the side information is used by the post-processor to assemble the reconstructed version X’ of the original document.Decomposition approaches yielding the same reconstructed image This figure describes the decomposition process. The basic approaches are:region classification (RC) and transition identification (TI). In case of RC decomposition, the regions containing graphics and text arerepresented in a separate FG plane. Everything is represented in FG plane including the spaces in between letters andother blank spaces. The mask as shown is uniform with large patches, clearly differentiating the textand the graphic regions and the background contains the document background,complex graphics and/or continuous tone pictures [1]. TI decomposition is quite similar to RC decomposition as can be seen from thesame figure. However, in case of TI, mask and FG planes represent graphics andtext in different manner. The FG plane pours the ink of text and graphics through the mask onto the BGplane. So the mask should have the necessary text contours. Hence the masklayer contains text characters, line art and filled regions, while the FG layercontains colors of the text letters and graphics. In RC decomposition, the mask layer is very uniform and very well compressible. But the FG can contain edges and continuous tone details. So it cannot becompressed well with typical continuous tone coders such as JPEG. As the mask layer in case of TI contains text objects and edges, it can be efficientlyencoded using standard binary coders such as JBIG and JBIG-2. The FG plane can be very efficiently coded even with coders such as JPEG becauseit contains large uniform patches. Also the FG plane can be sub-sampled withoutmuch loss in image quality [1].RD plot modification in multiple MRC layers[1]Fig. 3 RD plot modificationRD plot modification The benefit of using MRC model for compression can be observed by analyzing itsrate-distortion (RD) characteristics. As shown in Fig. 3, if image (a) is compressed with a generic coder A with fixedparameters except for a compression parameter, it will operate under a RD plot asshown in (b). It is seen that another coder B under the same circumstances isfound to perform better than coder A if its RD plot is shifted to the left as shown in(c). The logic for MRC is to split the image into multiple planes as shown in (d), and toapply coders C, D and E to each plane with RD plots similar to that of coder B.Thus the equivalent coder may have better RD plots than A, but there would be anoverhead associated with a multi-plane representation.List of existing MRC based encoders [2]RER(Resolution Enhanced Rendering)[2] In RER, Adaptive Error Diffusion method is used to encode edge detail into thebinary mask layer of the MRC document. The MRC document is then decoded using a Nonlinear Predictor to determine therelative amount of foreground and background color to be applied to each pixel. To yield minimum document image distortion, a method for jointly optimizing theparameters of the RER encoder and decoder is proposed. Simulation results indicate that RER method can reduce document imagedistortion to a great extent at a fixed bit rate. Also RER method is totally compatible with MRC standard and can be efficientlyimplemented in standard MRC encoders and decoders.Training model of the optimized encoder and decoderFig. 4: Comparison of MRC (a) and RER(Resolution Enhanced Rendering) (b) Encoders[2]XsXsRER Encoder In figure 4(b), the RER encoding module creates the dithered mask (Ds) as theoutput by taking in the FG, BG, binary mask layer, and the original document as theinput. The Ds, FG and BG are separately compressed by using the binary imageencoder and continuous-tone encoder that are used for the MRC encoder. In the encoding module, an edge detection procedure is performed on the binarymask layer to determine the pixels that lie on the boundary between FG and BG,to compute dithered
View Full Document
Unlocking...