Restoration of the Image Signal Lost by Analog to Digital Conversion

点击翻译按钮获取中文摘要

Resolution and bit-length of digital images are decided by Analog to Digital Conversion. In the case of video signals, frame rate is also decided by video recoding. Image enragement, frame rate up-conversion and bit-length expansion techniques are necessary for many digital image/video equipments, such as high-quality display such as the LCD display and the plasma display. In this research, three techniques(i.e., image enragement, frame rate up-conversion and bit-length expansion) are studied.Image enragement : The Laplacian pyramid representation(LP) method, which has been proposed by ourselves, is implemented on FPGA(Xilinx Virtex-II XC2V6000). The LP method is realized by 5654 slices, 7715 FFs and 42 block RAMs. We achieve 206.9[MHz] for maximum operating frequency and only 20 lines for system latency.Frame rate up-conversion : In this research, a novel frame rate up-conversion(FRUC) method using the motion compensation(MC) based on block matching algorithm(BMA) is proposed. The pro … More perty of the BMA is changed by the size of block. It is desirable for small moving object to set block size small. On the other hand, we would like to set the block size large in the global motion region. Therefore, the block size is decided depend on the property of the local region of frame image. A novel FRUC technique based on the motion compensated interpolation with multi-size blocks is proposed. At first, the frame image is divided into 8x8 blocks and motion vectors are estimated for each 8x8 block. Next, the similar motion blocks are merged into one block, up to 32x32 block. Thus the proposed method reduces block artifacts and realizes the clear interpolation of moving objects regardless of the size of those.Bit-length expansion : In this research, we propose a bit-length expansion method from M gray levels to N(=3(M-1) +1) gray levels. This operation means that two gray-levels are introduced between existing levels. The bit-length expansion method results in the estimation problem of an error image between an M gray level's image and an N gray level's image. The error image is constructed by only three values. If the difference between the existing level and the inserted new level is defined as Δ, the N level image is obtained by adding+Δ, 0 or-Δ to the M level image. We show that the proposed method can remove pseudo-contour of digital images through a lot of application results. Less