Title Page
ABSTRACT
Contents
I. Introduction 11
II. Background 13
2.1. Video Encryption 13
2.2. Conditional Access Control for Encrypted Scalable Contents 14
2.3. Scalable Video Coding(SVC) 16
III. Proposed Method 20
3.1. Requirements for SVC bitstream Encryption 20
3.2. Proposed SVC bitstream Encryption Algorithm 21
3.3. Conditional Access Control for Encrypted SVC bitstream 24
3.4. Key Management Scheme for Conditional Access Control 27
IV. Experiment Result 32
4.1. SVC bitstream Encryption 32
4.1.1. Encryption for the data type 32
4.1.2. Bitstream extraction after Encryption 35
4.2. Conditional Access Control experiments 40
V. Conclusion and Future Work 44
국문요약 46
References 50
Acknowledgement 52
Table 3.1. The NAL unit key set list needed to access SVC bitstream with certain scalability. It is calculated by Eq. (2) from the SVC example of Fig. 3.2. 27
Table 3.2. Access keys by the proposed algorithm from example of Fig. 3.2. 30
Table 3.3. NAL unit key set generated by the access key from the SVC example of Fig. 3.2. 31
Table 4.1. PSNR of decoded video for texture encryption and texture + motion vector encryption when base-layers are not encrypted 35
Table 4.2 shows the PSNR of decoded results for extraction in Fig. 4.3. Table 4.2 PSNR results for each extracted cases 37
Table 4.3. PSNR results for each FGS extraction case 39
Table 4.4 Access condition and corresponding key set to access 40
Table 4.5. PSNR results for access right quality and forced access quality 42
Fig. 2.1. Frequency band division by wavelet coding in JPEG 2000 15
Fig. 2.2. Access control for MPEG-4 FGS supporting PSNR and bitrate 16
Fig. 2.3. Spatial, Temporal and SNR, enhancement structure in SVC bitstream 17
Fig. 2.4. NAL unit syntax in SVC bitstream 18
Fig. 2.5. Bitstream extraction for the SVC bitstream, (a) NAL unit dropping for spatial and temporal scalability, (b) NAL unit cropping for SNR scalability 19
Fig. 3.1. Proposed encryption scheme for one layer in SVC 23
Fig. 3.2. NAL unit encryption for conditional access control 25
Fig. 3.3. Redefined two scalability types to categorize the type key 29
Fig. 4.1. Visual patterns of decoded video corrupted by no decryption key for respective texture, motion vector, and FGS encrypted videos 33
Fig. 4.2. Visual patterns of decoded video corrupted by no decryption key for respective texture and texture+motion vector encrypted videos when base-layers are not encrypted 34
Fig. 4.3. Visual patterns of decoded video corrupted by no decryption key after bitstream extraction by dropping NAL units 36
Fig. 4.4. SNR enhancement layers which consist of FGS data could be extracted by cropping NAL unit. (a), (c), (e) is the cases that FGS data is encrypted. (b), (d), (e) is the cases 38
Fig. 4.5. Visual patterns of decoded video for the given bitstream with access key of corresponding access right 41
Fig. 4.6. Comparison between the visual qualities for correct access right and wrong access right 43