Multi-modal image registration can provide richer and more comprehensive information than single-modal image registration. Among them, infrared and visible image registration, which is a common multi-modal form of registration, has important application value in fields such as electric power, remote sensing, military, and face recognition. In this paper, the correlation technique of infrared and visible image registration is introduced, and the existing difficulties and challenges involved in registration are analyzed. Subsequently, the advantages and disadvantages of different registration methods are evaluated in detail the three types based on area, feature, and deep learning, and a practical application of infrared and visible image registration technology is presented. Finally, the future development trend of infrared and visible image registration is discussed.
The color purity of organic light-emitting devices is restricted mostly by the intrinsic character of the emitting material; however, the optical microcavity can improve the color purity by changing the structure of the devices. In this study, we demonstrated that a high color purity monochromatic top-emitting device can be obtained by changing the length of the microcavity. By adjusting the thicknesses of the hole transport layer and electron barrier layer, high color purity organic green phosphorescent top-emitting electroluminescent devices were fabricated. The structure of the devices was an Si substrate/Ag/ITO/NPB:F16CuPc (10 nm, 3%)/NPB (x nm)/TCTA (y nm)/mCP:Ir(ppy)3 (40 nm, 6%)/Bphen:Liq (30 nm, 40%)/Mg:Ag (12 nm, 10%)/Alq3 (35 nm). Direct green emission with chromaticity coordinates of (0.2092, 0.7167) was obtained by changing the thicknesses of NPB and TCTA; this resulted in standard green light (0.21, 0.71).
Research on infrared small targets is crucial in the areas of military guidance and early warning and detection of border spy UAVs. In this paper, a small target tracking algorithm based on super-resolution enhancement and online detection DSST is proposed for small target tracking research. First, the original image is updated based on the integrated infrared image features of the super-resolution reconstruction algorithm to enhance the dim target. In addition, the enhanced image is used as the input for the online detection DSST algorithm to perform response mapping and estimate the target position. The experimental results show that the accuracy of the proposed algorithm is high compared with those of several new algorithms.
Infrared dim and small target (IRDST) detection is a longstanding and challenging problem in infrared search and track systems. To address the problems of a low detection rate and high false alarm rate for dim and small targets in complex backgrounds, a method is proposed for detecting IRDSTs using a regional bi-neighborhood saliency map (RBNSM). First, using the local a-priori property of the weak target, a sliding window is defined and divided into multiple cells before the mean value of the first maximum gray levels of the central cell is calculated to highlight the weak target. Then, the adjacent and spaced neighbors of the central cell are constructed and the mean value of their respective gray levels is calculated. Subsequently, the salient maps of the two neighbors are the extracted from different directions and multiplied point by point to further suppress the clutter background and enhance the weak target. Finally, the target is accurately detected by adaptive extraction. The detection results of various typical IR complex background images and SIRST datasets show that RBNSM has a better detection performance and clutter suppression ability in complex backgrounds than the seven representative methods.
Spectral graph wavelet transform (SGWT) can fully utilize the spectral characteristics of an image in the image domain and has advantages in the expression of small irregular regions. Therefore, this paper proposes an infrared and visible fusion algorithm based on multi saliency. First, SGWT is used to decompose the source image into a low-frequency sub-band and several high-frequency sub-bands. For low-frequency coefficients, a multi saliency fusion rule suitable for human visual features is proposed by combining multiple complementary low-level features. For high-frequency coefficients, a rule for increasing the absolute value of the region is proposed by fully considering the correlation of neighborhood pixels. Finally, a weighted least squares optimization method is applied to optimize the fusion image reconstructed by spectral wavelet reconstruction, which highlights the main target and retains the background details of visible light as much as possible. The experimental results show that, compared with seven related algorithms such as DWT and NSCT, this method can highlight the infrared target and retain more visible background details, resulting in a better visual effect. Moreover, it exhibits advantages in four objective evaluations: variance, entropy, Qabf, and mutual information.
Recently, multi-scale feature extraction has been widely used in the field of infrared and visible image fusion; however, most extraction processes are too complex, and the visual effect is not good. To improve the visual fidelity of the fusion result, a multi-scale horizontal image fusion model based on an edge-aware smoothing-sharpening filter (EASSF) is proposed for infrared and visible images. First, to obtain multiscale texture components and basic components in the horizontal direction, a multi-scale horizontal image decomposition method based on the EASSF is proposed to decompose the source image. Second, the maximum fusion rule is used to merge texture components, which can avoid loss of information detail. Then, to capture salient target information, the basic components are fused via the perceptual-fusion rule. Finally, the fused image is obtained by integrating the fused multi-scale texture components and basic components. By analyzing the perceptual fusion coefficient of PF, the appropriate range of infrared and visible image fusion in the multi-scale EASSF is obtained through the objective data of the fusion results. In this range, compared with several classical and popular fusion methods, the proposed fusion model not only avoids the complexity of feature information extraction, but also effectively ensures the visual fidelity of fusion results by integrating the significant spectral information of basic components.
To overcome the shortcomings of current infrared and visible image fusion algorithms, such as non-prominent targets and the loss of many textural details, a novel infrared and visible image fusion algorithm based on Gaussian fuzzy logic and the adaptive dual-channel spiking cortical model (ADCSCM) is proposed in this paper. First, the source infrared and visible images are decomposed into low- and high-frequency parts by non-subsampled shearlet transform (NSST). Then, these are combined with the new sum of the Laplacian and Gaussian fuzzy logic, and dual thresholds are set to guide the fusion of the low-frequency part; simultaneously, the fusion rule based on the ADCSCM is used to guide the fusion of the high-frequency part. Finally, the fused low- and high-frequency parts are reconstructed using inverse NSST to obtain the fused image. The experimental results show that the proposed algorithm has the best subjective visual effect and is better than the other seven fusion algorithms in terms of mutual information, information entropy, and standard deviation. Furthermore, the proposed algorithm can effectively highlight the infrared target, retain more textural details, and improve the quality of the fused image.
To address the problems of low diopter recognition accuracy and low detection efficiency in the pupil area, this paper proposes a pupil image detection algorithm based on an improved YOLOv3 deep neural network. First, a two-class detection network YOLOv3 base for extracting the main features of the pupil is constructed to strengthen the learning ability of the pupil characteristics. Subsequently, through migration learning, the training model parameters are migrated to YOLOv3-DPDC to reduce the difficulty of model training and poor detection performance caused by the uneven distribution of sample data. Finally, fine-tuning is used to quickly train the YOLOv3 multi-classification network to achieve accurate pupil diopter detection. An experimental test was performed using the 1200 collected infrared pupil images. The results show that the average accuracy of diopter detection using this algorithm is as high as 91.6%, and the detection speed can reach 45 fps; these values are significantly better than those obtained using Faster R-CNN for diopter detection.
This paper presents a pulse-coupled neural network (PCNN) method for infrared fault region extraction based on maximum similarity thresholding to detect the fault region from the infrared image of a transmission line. In this method, the synchronous pulse characteristics of the PCNN model are used to cluster pixels via inner iteration, and the model is simplified by incorporating the maximum similarity thresholding method, enabling the PCNN model to simplify the thresholding setting. Meanwhile, the minimum clustering variance is introduced to set the linking coefficient. Thus, the PCNN model can efficiently segment an infrared image and obtain the effective thermal fault region in the image. The experimental results show that the proposed method exhibits good performance in region extraction and may be suitable for increasing the efficiency of automatic fault detection along transmission lines.
To solve problems in traditional image fusion, such as dim targets, low contrast, and loss of edge and textural details in fusion results, a new fusion approach for infrared and low-level visible light image fusion based on perception unified color space (PUCS) and dual tree complex wavelet transform (DTCWT) is proposed. First, the two-source image intensity component is separately transformed from RGB space into PUCS to obtain a new intensity component for further processing. Then, the infrared and low-level visible light images are decomposed using DTCWT to obtain the low- and high-frequency components, respectively. Subsequently, at the fusion stage, the region energy adaptive weighted method is adopted to fuse the low-frequency sub-bands, and the high-frequency rule uses the sum modified Laplacian and gradient value vector for different scale and directional sub-bands fusions. Finally, the fusion image is obtained by applying inverse DTCWT on the sub-bands and returned to RGB space. The proposed algorithm was compared with three efficient fusion methods in different scenarios. The experimental results show that this approach can achieve prominent target characteristics, clear background texture and edge details, and suitable contrast in subjective evaluations as well as advantages in eight objective indicator evaluations.
To address the problems of image fusion in the spatial domain, such as the extraction of different image sources, and challenges in selecting fusion weights, a new spatial-domain image-fusion algorithm is proposed. Using the basic principle of matrix similarity, the infrared image matrix is diagonally transformed and the visible light image matrix is mapped onto the main eigenvectors. Then, the weighted fusion method is used to process the eigenvalue matrix and the fusion matrix is diagonalized as an inverse-transformed and reconstructed fusion image. The experimental results show that the algorithm fully retains the effective information of the source image; moreover, the overall grayscale of the fused image is significantly improved. Thus, the algorithm offers a strong image quality evaluation index and better visual effects.
A two-axis horizontal frame coarse tracking was designed using the application requirements of space debris capture systems. A PI speed loop control method based on closed-loop acceleration was proposed to realize closed-loop high bandwidth control and high-precision tracking accuracy. First, the horizontal coarse tracking frame was designed based on the beam propagation path and geometric load size requirements. The model of the single-axis structure was simplified, and a dynamic equation for the simplified model of the damping stiffness of the two-dimensional single axis was established. Subsequently, vibration analysis was performed to determine the resonance frequency, locked rotation frequency, and main structural parameters of the tracking frame. Third, a double closed-loop control system with velocity and acceleration feedback was designed, and the parameters of the control system were determined. Finally, a performance test of the control system was conducted. The results showed that the control system meets the performance demands. The bandwidth of the control system was 28.2% greater than that of the PI speed loop control system. The PI speed loop control system based on closed-loop acceleration improved the adjustment time by 78.6% and reduced the overshoot by 94.08%. The PI position loop control system based on closed-loop acceleration had an adjustment time of 0.085 s and an overshot of 11.66%, which exhibited a small tracking error and strong antiinterference ability.
The identification of cracks in the metal structure of lifting machinery is a new direction for infrared thermal imaging detection technology. In this study, the detection principle of pulsed infrared thermal imaging was introduced, and a pulsed infrared thermal imaging detection system was designed; the experimental platform was constructed on the basis of these. Median filtering and Butterworth low-pass filtering were used to process the images collected in the experiment. To address the problem of blurring at the edges of the defects after processing the above algorithms, a Butterworth band-pass filtering algorithm was proposed. After threshold segmentation and edge detection, the defect contour feature was extracted, and using the conversion relationship between the actual size of the flat specimen and the contour feature image pixels, the recognition accuracy of the crack defect was finally obtained. The comparison and verification demonstrated that the pulsed infrared thermal imaging technology can meet the requirements of crack defect detection in crane metal structures.
To address the difficulty in detecting the source of roof leakage, an image enhancement method that uses the infrared image features of the leakage area was studied using gray segmentation mapping. Rapid image recognition technology based on a template matrix was proposed, and an automatic roof leakage source detection system was designed. Leakage sources were set on a 5 mx 3 m roof to form multiple leakage areas. A mecanum wheeled trolley was used to support the system while detecting these sources. The results showed that the system could complete detection within 89 s, with a total of 150 leakage points tested and 12 leakage points missed, and the identification accuracy was greater than 90%. This technology has high detection efficiency and simple operation and can be used to detect all types of unknown water seepage sources with the corresponding carrier.
This article introduces the manufacturing process, material selection, sealing characteristics, and sealing advantages of a new hollow metal C ring used in Stirling cryocoolers. Given the high leakage rate and ease of failure of rings made of traditional rubber, PTFE, and solid silver wire, the manufacturing accuracy of the metal C-type sealing ring is high. After years of testing, use, and demonstrations, the metal C ring has achieved high reliability and a low leakage rate under conditions of strong impact and radiation resistance. The leakage rate of the metal C-ring can reach to 10-9 atm.cm3/s.m, which meets the leakage rate requirements of Stirling cryocoolers.