Infrared Image Super Resolution by Combining Compressive Sensing and Deep Learning

Abstract

Super resolution methods alleviate the high cost and high difficulty in applying high resolution infrared image sensors. In this paper we present a novel single image super resolution method for infrared images by combining compressive sensing theory and deep learning. Low resolution images can be regarded as the compressed sampling results of the high resolution ones in compressive sensing. With sparsity in this theory, higher resolution images can be reconstructed. However, because of diverse level of sparsity for different images, the output contains noise and loss of high frequency information. Deep convolutional neural network provides a solution to relieve the noise and supplement some missing high frequency information. By concatenating two methods, we manage to produce better results in super resolution tasks for infrared images than SRCNN and ScSR. PSNR and SSIM values are used to quantify the performance. Applying our method to open datasets and actual infrared imaging experiments, we also find better visual results are preserved.

Thermal Image Reconstruction Using Deep Learning

“…For the comparison, the experiment is performed using six open datasets as shown in Figure 6(i), (j), (k), (l), (m, n, o), and (p). The results of SRGAN [17] and the proposed SRR method in Table 15 are obtained in this study, and the results of the remaining methods, such as SRCNN [18], ScSR [88], and Zhang et al's method [33], are included. Furthermore, the results in Tables 16 and 17 are obtained through an experiment in this study.…”

“…Table 15 compares the methods using the upscaling factor of 2. Moreover, no experiment was performed with the upscaling factor set to 4 in the previous study [33]. However, as most of the existing SRR-based methods are compared based on the performance using the upscaling factor of 4, the proposed method was compared with the performances of Bicubic [89], SRCNN, SRGAN, PULSE [90] and by setting the upscaling factor to 4, as shown in Table 16.…”

“…Ref. [33] proposed a method of performing SRR for thermal images by combining compressive sensing theory and deep learning. In ref.…”

“…With deeplearning method 1-channel image [33][34][35] The method extracts suitable features in various camera settings and environments.…”

Infrared Image Super-Resolution Reconstruction Based on Quaternion and High-Order Overlapping Group Sparse Total Variation

“…In fact, for the noisy images, the conventional super-resolution way is to denoise the images as a pre-processing step and then super-resolve the denoised images. In some new methods [46,47,48,49], such as the median filter transform (MFT) with parallelogram-shaped windows [47], denoising and super-resolving are integrated to provide improved results in comparison to the conventional way.…”