Cruciate Block Compression Foveated JND Model for Near‑Eye Display

Yanrui Sun; Yuan Ji

doi:10.3788/CJL231552

Journals >Chinese Journal of Lasers >Volume 51 >Issue 17 >Page 1709001 > Article

Chinese Journal of Lasers
Vol. 51, Issue 17, 1709001 (2024)

Cruciate Block Compression Foveated JND Model for Near‑Eye Display

Yanrui Sun¹ and Yuan Ji^1,2,*

Author Affiliations

¹Microelectronics Research and Development Center, Shanghai University, Shanghai 200072, China

²School of Mechatronic Engineering and Automation, Shanghai University, Shanghai 200072, China

show less

DOI: 10.3788/CJL231552 Cite this Article Set citation alerts

Yanrui Sun, Yuan Ji. Cruciate Block Compression Foveated JND Model for Near‑Eye Display[J]. Chinese Journal of Lasers, 2024, 51(17): 1709001 Copy Citation Text

show less

Fig. 1. Luminance masking test image

Download full size

Fig. 2. Contrast masking test image

Download full size

Fig. 3. Results of visual experiment. (a) JND-luminance; (b) JND-eccentricity; (c) JND-contrast; (d) JND-block size

Download full size

Fig. 4. Relationship among JND and different parameters. (a) Relationship among JND, luminance and eccentricity; (b) relationship among JND, luminance and contrast; (c) relationship among JND, block size and eccentricity

Download full size

Fig. 5. CB-FJND model noise injection. （a） Original image；（b） noise injection with 3 pixel×3 pixel block；（c） noise injection with 5 pixel×5 pixel block；（d） noise injection with 7 pixel×7 pixel block；（e） result of Ref. ［7］ model；（f） results of Ref. ［9］ model

Download full size

Fig. 6. Lena bit-plane images. (a) 23 bit-planes; (b) 25 bit-planes

Download full size

Fig. 7. Distribution of bit16 before and after modification of

Download full size

Fig. 8. Comparison of compression effect and partially enlarged image. (a) Algorithm in Ref. [10]; (b) cruciate block compression algorithm (19 subfields); (c) cruciate block compression algorithm (25 subfields)

Download full size

Fig. 9. Hardware implementation system block diagram

Download full size

Fig. 10. Picture of FPGA test platform

Download full size

Fig. 11. Image compression effect and detail enlargement comparison of Lena. (a) Algorithm in Ref. [10]; (b) our algorithm (19 subfields); (c) our algorithm (25 subfields)

Download full size

Attribute	Luminance masking test	Contrast masking test
Eccentricity /（°）	1，2，3，5，10，15，20，25，30，35，40，45，50	1，2，3，5，10，15，20，25，30，35，40，45，50
Luminance	0，10，20，40，60，80，100，125，150，180，200，220，240	60，100，150，180，220
Contrast	0	0，20，40，60，80

Table 1. Luminance/contrast masking test variable control

Attribute	Value
Block size /pixel	2，3，5，7，10
Luminance	0，10，20，40，60，80，100，125，150，180，200，220，240
Eccentricity /（°）	1，2，3，5，10，15，20，25，30，35，40，45，50

Table 2. Block masking test variable control

Stage	Maximum run length	Average rate of variability /%
Before modification	49	37.86
After modification	61	29.88

Table 3. Comparison of bit16 before and after threshold modification

Subfield number	Weight distribution
19	16｛15｝ 8 4 2 1
20	16｛15｝ 8 4 1｛3｝
21	16｛15｝ 8 2｛2｝ 1｛3｝
22	16｛15｝ 8 2 1｛5｝
23	16｛15｝ 8 1｛7｝
24	16｛15｝ 4｛2｝ 1｛7｝
25	16｛15｝ 4 2｛2｝ 1｛7｝

Table 4. 19‒25 bit-plane weights

Subfield number	Fractional subfield weights
20	8∶4∶1∶1∶1
	8∶2∶2∶2∶1
	4∶4∶4∶2∶1
21	8∶2∶2∶1∶1∶1
	4∶4∶4∶1∶1∶1
	4∶4∶2∶2∶2∶1
22	8∶2∶1∶1∶1∶1∶1
	4∶4∶2∶2∶1∶1∶1
	4∶2∶2∶2∶2∶2∶1
23	8∶1∶1∶1∶1∶1∶1∶1
	4∶4∶2∶1∶1∶1∶1∶1
	4∶2∶2∶2∶2∶1∶1∶1
	2∶2∶2∶2∶2∶2∶2∶1
24	4∶4∶1∶1∶1∶1∶1∶1∶1
	4∶2∶2∶2∶1∶1∶1∶1∶1
	2∶2∶2∶2∶2∶2∶1∶1∶1
25	4∶2∶2∶1∶1∶1∶1∶1∶1∶1
25	2∶2∶2∶2∶2∶1∶1∶1∶1∶1

Table 5. Fractional subfield weight division strategy

Subfield number	Fractional subfield weights	PSNR /dB			SSIM
Subfield number	Fractional subfield weights	Lena	Peppers	Baboon	Lena	Peppers	Baboon
20	8∶4∶1∶1∶1	43.2422	42.7886	41.8702	0.9988	0.9987	0.9968
	8∶2∶2∶2∶1	41.1474	40.4945	38.9426	0.9979	0.9980	0.9959
	4∶4∶4∶2∶1	38.0684	37.6907	37.7089	0.9959	0.9964	0.9922
21	8∶2∶2∶1∶1∶1	41.1474	40.4945	38.9426	0.9979	0.9980	0.9959
	4∶4∶4∶1∶1∶1	38.2299	37.9056	37.7089	0.9961	0.9966	0.9922
	4∶4∶2∶2∶2∶1	38.1131	37.7037	37.7089	0.9959	0.9965	0.9922
22	8∶2∶1∶1∶1∶1∶1	40.5129	39.9217	38.9426	0.9977	0.9976	0.9959
	4∶4∶2∶2∶1∶1∶1	38.1131	37.7037	37.7089	0.9959	0.9965	0.9922
	4∶2∶2∶2∶2∶2∶1	35.4802	36.2174	34.9804	0.9920	0.9949	0.9907
23	8∶1∶1∶1∶1∶1∶1∶1	38.7953	38.3462	37.7445	0.9967	0.9962	0.9953
	4∶4∶2∶1∶1∶1∶1∶1	37.8025	37.6806	37.7089	0.9956	0.9965	0.9922
	4∶2∶2∶2∶2∶1∶1∶1	35.4802	36.2174	34.9804	0.9920	0.9949	0.9907
	2∶2∶2∶2∶2∶2∶2∶1	32.4729	32.7110	34.3232	0.9834	0.9909	0.9864
24	4∶4∶1∶1∶1∶1∶1∶1∶1	37.7469	37.8498	37.7089	0.9955	0.9966	0.9922
	4∶2∶2∶2∶1∶1∶1∶1∶1	35.4802	36.2174	34.9804	0.9920	0.9949	0.9907
	2∶2∶2∶2∶2∶2∶1∶1∶1	32.4729	32.7110	34.3232	0.9834	0.9909	0.9864
25	4∶2∶2∶1∶1∶1∶1∶1∶1∶1	35.4802	36.2174	34.9804	0.9920	0.9949	0.9907
25	2∶2∶2∶2∶2∶1∶1∶1∶1∶1	32.4729	32.7110	34.3232	0.9834	0.9909	0.9864
JPEG quality 100		37.4874	32.8036	29.2941	0.9946	0.9886	0.9511
JPEG quality 90		34.7660	31.3296	27.9459	0.9902	0.9826	0.9333
JPEG quality 80		33.6061	30.5420	26.6553	0.9875	0.9789	0.9143

Table 6. Image quality after compression with different weight distribution bit-planes

Image	Ref. ［10］	Ref. ［11］	CB-FJND with different subfield numbers
Image	Ref. ［10］	Ref. ［11］	19	20	21	22	23	24	25
Lena	36.00	53.47	36.57	32.96	31.51	30.26	28.71	27.82	26.10
Peppers	40.20	56.22	40.18	36.42	34.81	33.39	31.74	30.94	29.10
Baboon	51.03	63.68	50.43	47.67	45.06	43.15	41.31	39.91	36.84
Average	42.41	57.79	42.39	39.02	37.13	35.6	33.92	32.89	30.68

Table 7. Bit compression rate of compressed image

Image	Ref. ［10］	Ref. ［11］	CB-FJND with different subfield numbers
Image	Ref. ［10］	Ref. ［11］	19	20	21	22	23	24	25
Lena	38.92	37.73	43.63	43.24	41.15	40.52	38.80	37.75	35.48
Peppers	39.85	38.16	43.44	42.79	40.50	39.92	38.35	37.85	36.21
Baboon	37.44	36.91	42.81	41.87	38.95	38.95	37.75	37.71	34.98
Average	38.73	37.60	43.29	42.63	40.20	39.80	38.30	37.77	35.56

Table 8. PSNR of compressed image

Image	Ref. ［10］	Ref. ［11］	CB-FJND with different subfield numbers
Image	Ref. ［10］	Ref. ［11］	19	20	21	22	23	24	25
Lena	0.9974	0.9813	0.9989	0.9988	0.9979	0.9977	0.9967	0.9955	0.9920
Peppers	0.9973	0.9807	0.9988	0.9987	0.9980	0.9976	0.9962	0.9966	0.9949
Baboon	0.9955	0.9792	0.9975	0.9968	0.9959	0.9959	0.9953	0.9922	0.9907
Average	0.9967	0.9804	0.9984	0.9981	0.9973	0.9971	0.9961	0.9948	0.9925