Enhanced AC Coefficients Compression Techniques

Explore improved methods for compressing AC coefficients in JPEG coding, utilizing alternate adaptive scanning and multi-feature determined context arithmetic coding. The approach involves scanning order determination, entropy coding, and arithmetic coding techniques to enhance compression efficiency.

• Compression Techniques
• JPEG Coding
• Adaptive Scanning
• Arithmetic Coding
• Entropy Coding

cal_jo Follow

Uploaded on Dec 08, 2024 | 0 Views

Download Presentation

Please find below an Image/Link to download the presentation.

The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.

Presentation Transcript

1. Improved AC Coefficients Compression Using Alternate Adaptive Scanning and Multi-Feature Determined Context Arithmetic Coding Presenter: HSU, HUANG-CHUN 1

2. JPEG Coding Zig-zag scan Alternative scan L:G are coded together Coded separately Huffman code CAAC Additional Bit Length Index Nonzero AC Values 33 3 2 0 0 0 0 0 5 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 2 2 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 -1, 1 1 1 0 0 0 0 0 0 2 -3, -2, 2, 3 2 3 -7, , -4, 4, , 7 3 4 -15, , -8, 8, , 15 4 5 -31, , -16, 16, , 31 5 6 -63, , -32, 32, , 63 6 7 -127, , -64, 64, , 127 7 8 -255, , -128, 128, , 255 8 5, 3,2,0,0,0,0,0,0,0,0,2, 1, 2, 1, 2,0, ,0 9 -511, , -256, 256, , 511 9 10 -1023, , -512, 512, , 1023 10 0: 5 , 0: 3 , 0:2 , 8: 2 , 0: 1 , 0: 2 , 0: 1 , 0: 2 ,EOB Huffman code for AC coefficients [1] 2

3. Scanning Order Determination Zig-zag term should be conserved (ZZ) but without bias Probability of zero in neighboring blocks (PZ) Neighbor is defined as below Gradient by DC terms (GD) based on the postulate that DC is encoded prior to AC 3 ZZ Neighbor DC

4. Scanning Order Determination Contd 1 1 1 Some parameters will be set to the power of PZ, GD, PR 4

5. Entropy Coding Separated Coding L and G are negatively correlated 3 Contexts are assigned Therefore L is encoded prior to G 5 Context 1 determination An example of L:G pairs histogram

6. Entropy Coding Contd Context 2 s value is either 1 or 0 if exceeds the corresponding element o of threshold vector [5, 3.5, 2.3, 1.8, 1.35, 1.1, 1] 6 The average absolute values of nonzero coefficients for the quantized 8x8 DCT block [2]

7. Entropy Coding Contd 1 Total number of contexts will be 56 7

8. Arithmetic Coding Techniques Frequency table initialization Hyper-Laplacian form Range increment 1 Adaptive bounding similar to CAVLC Long-term inputs suppression 1 8

9. Results Length of bitstream and the corresponding bpp reduction of AC coefficients with larger input size (1024*1024) 9 Length of bitstream and the corresponding bpp reduction of AC coefficients with smaller input size (Mostly 512*512)

10. Reference [1] ITU-T T.81 (1992) | ISO/IEC 10918-1:1993, Information technology Digital compression and coding of continuous-tone still images Requirements and guidelines. [2] Y. S. Lu, Context Assignment Algorithms for Image Compression. National Taiwan University, Graduate Institute of Communication Engineering master thesis, Taipei (2021) 10