저복잡도 IDCT를 위한 제로 계수 인식 IDCT 알고리즘

Abstract

Ever since many well-known image and video coding standards such as JPEG, MPEG, and H.26x started using discrete cosine transform (DCT) as a core process in data compression, the design of the fast inverse DCT (IDCT) has been an intensive research topic. Most research has focused on a reduction of the number of operations using the butterfly structure. However, the majority of DCT coefficients is zero after quantization and is therefore redundant for IDCT computation. Therefore, we exploited this DCT coefficients redundancy to propose a zero coefficient-aware IDCT algorithms for fast decoding. The proposed zero coefficient-aware fast IDCT algorithms efficiently apply the context information: coefficient-level zero coefficient information and one-dimensional line-level zero line information. In the coefficient-level zero coefficient-aware fast IDCT algorithms, the proposed two methods significantly reduce the number of IDCT operations by adaptively including non-zero coefficients in the calculation and employing a table look-up to eliminate multiplication operations in the IDCT process. In the one-dimensional line-level zero coefficient-ware fast IDCT algorithm, the proposed method focuses on the reduction of the computational complexity of the 8×8 IDCT in the HEVC decoding process by removing the unnecessary computation of one-dimensional IDCTs as well as by adaptively applying IDCT kernels, based on the number of non-zero DCT coefficients. The proposed methods are to confirm the efficiency through the experimental results. The proposed coefficient-level zero coefficient-ware IDCT algorithm outperformed other existing fast IDCT algorithms in terms of operational complexity. The running time was faster than the conventional IDCT algorithms implemented in MPEG-4 simple profile decoder by a speedup factor of 1.32 times for the SIF/CIF sequences and up to 2.18 times for the high definition (HD) sequences. In addition, the proposed coefficient-level scaled zero coefficient-ware IDCT algorithm significantly improves the computation complexity, which applies the concept of the fixed-point approximation with the accurate dyadic terms to the coefficient-level zero coefficient-aware IDCT algorithm. The experimental result showed that the running time of the proposed scaled zero coefficient-ware algorithm was faster than that of the fixed-point 8x8 IDCT standard (ISO/IEC 23002-2) by a speedup factor of 2.0 times on an average and up to 2.6 times for the HD sequences. Experimental results for the line-level zero coefficient-ware algorithm showed that the average operation using the proposed IDCT kernel design was approximately half of the 8x8 IDCT in the HEVC test model decoder. The improved computational complexity of the proposed method was demonstrated by the running time, which required only half on average, of the total decoding time of IDCT module as compared to IDCT in HM decoder.