Infrared and Visible Image Fusion in Latent Low Rank Representation Framework Based on Convolution Neural Network and Guided Filtering

Xicheng LOU; Xin FENG

doi:10.3788/gzxb20215003.0310004

Journals >Acta Photonica Sinica >Volume 50 >Issue 3 >Page 180 > Article

Acta Photonica Sinica
Vol. 50, Issue 3, 180 (2021)

Infrared and Visible Image Fusion in Latent Low Rank Representation Framework Based on Convolution Neural Network and Guided Filtering

Xicheng LOU and Xin FENG

Author Affiliations

Key Laboratory of Manufacturing Equipment Mechanism Design and Control of Chongqing， College of Mechanical Engineering， Chongqing Technology and Business University， Chongqing400067， China

show less

DOI: 10.3788/gzxb20215003.0310004 Cite this Article

Xicheng LOU, Xin FENG. Infrared and Visible Image Fusion in Latent Low Rank Representation Framework Based on Convolution Neural Network and Guided Filtering[J]. Acta Photonica Sinica, 2021, 50(3): 180 Copy Citation Text

show less

Fig. 1. LatLRR decomposition results

Download full size | View in the Article

Fig. 2. Neuron structure

Download full size | View in the Article

Fig. 3. Siamese network structure

Download full size | View in the Article

Fig. 4. CNN processing diagram

Download full size | View in the Article

Fig. 5. GF processing diagram

Download full size | View in the Article

Fig. 6. The framework of proposed method

Download full size | View in the Article

Fig. 7. Comparison of the first set of infrared and visible image fusion results

Download full size | View in the Article

Fig. 8. Comparison of the second set of infrared and visible image fusion results

Download full size | View in the Article

Fig. 9. Comparison of the third set of infrared and visible image fusion results

Download full size | View in the Article

Fig. 10. Comparison of the fourth set of infrared and visible image fusion results

Download full size | View in the Article

Image	Fusion method	$Q_{M I}$	$Q_{P}$	$Q_{S}$	$Q_{C B}$
First set of fusion results	Proposed method	0.374 1	0.416 8	0.789 4	0.589 6
	Ref. [17]	0.285 0	0.324 4	0.796 1	0.582 4
	Ref. [15]	0.257 0	0.250 6	0.803 8	0.519 1
	SR	0.298 7	0.167 7	0.726 2	0.552 0
	DTCWT	0.223 1	0.273 1	0.790 8	0.549 8
	CVT	0.210 4	0.220 7	0.780 4	0.528 6
	NSCT_SR	0.232 2	0.251 7	0.784 2	0.564 0
Second set of fusion results	Proposed method	0.460 3	0.435 8	0.784 6	0.550 5
	Ref. [17]	0.347 5	0.477 0	0.810 7	0.522 6
	Ref. [15]	0.280 6	0.336 1	0.783 8	0.466 9
	SR	0.392 6	0.234 2	0.704 6	0.503 2
	DTCWT	0.237 8	0.428 7	0.804 0	0.456 9
	CVT	0.224 5	0.359 1	0.796 3	0.455 5
	NSCT_SR	0.307 1	0.383 5	0.781 3	0.462 2
Third set of fusion results	Proposed method	0.399 5	0.263 1	0.844 9	0.466 5
	Ref. [17]	0.335 2	0.255 7	0.824 6	0.486 2
	Ref. [15]	0.218 2	0.230 6	0.801 7	0.461 1
	SR	0.444 4	0.117 3	0.780 9	0.494 1
	DTCWT	0.185 6	0.212 7	0.845 5	0.487 8
	CVT	0.163 3	0.184 6	0.830 6	0.477 1
	NSCT_SR	0.498 3	0.256 6	0.842 2	0.520 6
Fourth set of fusion results	Proposed method	0.358 1	0.414 5	0.882 1	0.474 4
	Ref. [17]	0.378 5	0.509 6	0.764 2	0.553 1
	Ref. [15]	0.333 9	0.389 0	0.801 8	0.511 6
	SR	0.512 9	0.382 4	0.796 3	0.579 4
	DTCWT	0.502 4	0.331 5	0.788 0	0.548 3
	CVT	0.283 3	0.414 8	0.847 3	0.508 5
	NSCT_SR	0.466 4	0.435 5	0.870 9	0.564 1

Table 1. Objective evaluation results of different fusion methods

Download Citation

Save the article for my favorites

Paper Information