The rapid development of the Drosophila-like compound eye visual neural computational modeling technology has improved the real-time detection and tracking of motion targets and other bionic applications a research challenge. The Drosophila-like compound eye visual neural system is a special highly parallel visual neural network model that uses moderate complexity to provide fast visual information processing. In this study, the status of current research on the visual neural computational modeling of the Drosophila-like compound eye is discussed in detail in four parts: physiological structure of the Drosophila-like compound eye, photoelectric conversion of the photoreceptor cells, processing of the optic lobe nerve, and cognitive decision of the central brain. The typical bionic applications of the Drosophila-like compound eye visual neural computational modeling in the military and civilian fields are introduced, and the development trends and challenges of the computational modeling are considered.

This study proposes an improved sparrow search algorithm based on an infrared temperature-data-optimized back propagation neural network (ISSA-BPNN) for an airborne circuit board chip fault diagnosis method that cannot diagnose dynamic failures. First, an infrared thermal imaging camera collected circuit board chip temperature data to establish a feature model of static, dynamic, and statistical characteristics of the circuit board chip warming process. We used sine chaos mapping to initialize the sparrow population distribution, the levy flight improvement finder sparrow location update, and an improved sparrow search algorithm to optimize the weight parameters of the BP neural network. Finally, the temperature feature model was input to the ISSA-BP neural network for training and testing to complete the circuit board chip fault diagnosis. The experiments used an avionics system power supply circuit board for reliability analysis, and the results revealed that the method achieved a comprehensive fault diagnosis rate of 97.84% under different circuit board operating conditions.

To solve the problems of vague targets, detail loss, and algorithm instability in traditional infrared and visible-light image fusion algorithms, a fusion method based on fuzzy c-means (FCM) clustering and guided filtering is proposed. The low-frequency sub-band was enhanced by guided filtering after applying a non-subsampled shearlet transform (NSST) to the original image. The low-and high-frequency sub-bands were then fused using FCM clustering and a dual-channel spiking cortical model. Finally, the fused image was obtained using an inverse NSST transform. The experimental results showed that the proposed algorithm was stable, the fusion image had clear targets and relatively complete details in the subjective evaluation, and the algorithm had an excellent standard deviation, mutual information, average gradient, information entropy, and edge retention factor in the objective evaluation.

In specific environments, when an infrared sensor cannot detect a target, it is necessary to integrate polarization and infrared technologies. To obtain a clearer fused image, this study adopted a method based on a multiscale structure and feature image fusion to realize infrared and polarization image fusion. The algorithm decomposed the infrared image and polarization map into three independent parts: average intensity, signal intensity, and signal structure. An arctangent weight function was proposed for fusion in the average intensity part, the signal intensity adopted the maximum fusion principle, and the signal structure adopted a weighted average square based on the power function of the signal intensity for fusion, and finally, the fused image was reconstructed. To fuse faster and reduce computational complexity, the decomposition process was replaced with mean filtering, and the final fused image was obtained by upsampling and downsampling. To obtain a better fusion image, better fusion parameters were selected through an experimental comparison of different fusion parameters. Experiments showed that by using the proposed arctangent weight function and fusion parameter setting, the four evaluation indexes had advantages over the traditional multiscale algorithm and subjectively retained more texture details, improved contrast, and suppressed artifacts.

Mainstream fusion methods based on deep learning employ a convolutional operation to extract local image features; however, the interaction between an image and convolution kernel is content-independent, and the long-range dependency cannot be well modeled. Consequently, the loss of important contextual information may be unavoidable and further limit the fusion performance of infrared and visible images. To this end, we present a simple and effective fusion network for infrared and visible images, namely, the multiscale transformer fusion method (MsTFusion). We first designed a novel Conv Swin Transformer block to model long-range dependency. A convolutional layer was used to improve the representative ability of the global features. Subsequently, we constructed a multiscale self-attentional encoding-decoding network to extract and reconstruct global features without the help of local features. Moreover, we designed a learnable fusion layer for feature sequences that employed softmax operations to calculate the attention weight of the feature sequences and highlight the salient features of the source image. The proposed method is an end-to-end model that uses a fully attentional model to interact with image content and attention weights. We conducted a series of experiments on TNO and road scene datasets, and the experimental results demonstrated that the proposed MsTFusion transcended other methods in terms of subjective visual observations and objective indicator comparisons. By integrating the self-attention mechanism, our method built a fully attentional fusion model for infrared and visible image fusion and modeled the long-range dependency for global feature extraction and reconstruction to overcome the limitations of deep learning-based models. Compared with other state-of-the-art traditional and deep learning methods, MsTFusion achieved remarkable fusion performance with strong generalization ability and competitive computational efficiency.

A simulation method for the infrared three-dimensional (3D) image reconstruction of a high-voltage switchgear temperature field based on a lazy snapping hybrid simulated annealing algorithm is proposed to improve the simulation effect of the high-voltage switchgear temperature field. In this method, the active infrared image sensor was used to collect an infrared image of the temperature field of a high-voltage switchgear, and the square-graph nonlinear stretching algorithm was used to enhance it. Subsequently, the enhanced infrared image was segmented by lazy snapping algorithm, and the two-dimensional texture flow-field image was synthesized by a hybrid genetic simulated annealing algorithm. Based on the image, an image quilting texture synthesis algorithm was used to synthesize the 3D image of the high-voltage switchgear temperature field, and the infrared 3D image reconstruction simulation process of the temperature field was completed. The experimental results demonstrated that this method effectively enhanced the infrared image of the temperature field of a high-voltage switchgear, and the intersection and union ratio of the segmented infrared image had a value that was approximately 1.0. The 3D image was effectively generated according to the generated two-dimensional texture flow-field image, and the maximum deviation of the temperature field simulation value was only 0.03℃.

To improve the performance of visible and infrared image decision-making-level fusion target detection algorithms, a decision-level fusion strategy based on model reliability was proposed. First, image preprocessing technology was adopted to improve the overall quality of an infrared image, and then visible and thermal infrared target detection models were trained and tested. The parameters required for the fusion strategy were obtained based on the model test results. The model detection results were fused according to the proposed fusion strategy, and the final fusion detection results were obtained. The experimental results showed that compared with the detection results of a single-target detection model, the missed detection rate of the fusion algorithm used in the daytime was 8.16% lower than that of the visible detection model, and the missed detection rate at night was 9.85% lower than that of the infrared detection model.

To conduct a systematic spectral radiation characteristic test of ground targets, the accuracy of the spectral radiometer must be simultaneously determined. The principle of the ground target infrared radiation characteristic test, spectral data calibration, target test, and atmospheric transmittance test method based on a spectral radiometer are introduced in this study. Using the infrared radiation characteristic test of an aeroengine based on a spectral radiometer as an example, the test scheme design, implementation process, results, and analysis are described in detail. The experimental data demonstrated the feasibility and accuracy of the method described in this study. The infrared radiation characteristic test method for aeroengines can be applied to testing and evaluation of other types of ground targets and supporting research on the infrared spectral and stealth characteristics of ground targets.

To reduce the effect of vibration on camera image quality during flight under heavy optical loads, a new type of centripetal shock absorber was designed. In view of the shortcomings of traditional shock absorbers, a design scheme for the centripetal structure is proposed that can effectively decouple the force. By uniformly distributing the force along the ring on the lifting plate, collisions within the optical load were effectively avoided. To test the vibration-damping effect of the centripetal structure, the optical load was verified using a combination of tensile tests and finite element analysis. The results showed that under a load of 120 kg, the maximum deformation of the entire structure was 3 mm, the maximum stress was 34.3 MPa, and the actual mode frequency was 9.81 Hz, which met the design requirements.

We prepared top-emitting white organic light-emitting diode (OLED) devices using Alq3 and ZnSe as capping layers (CPLs). The basic structure of the OLED was ITO/NPB:LiQ(5%)(10 nm)/TCTA(20 nm)/FIrpic+3.5%Ir(ppy)3+0.5%Ir (MDQ)2(acac)(25 nm)/TPBI(10 nm)/LiF(5 nm)/Mg:Ag(10%)(12 nm)/ CPL. The brightness and current efficiency of OLED devices fabricated with Alq3 and ZnSe as CPLs are significantly better than those without capping layers. Moreover, the color coordinates (CIEX, CIEY) of the OLED devices prepared with ZnSe CPLs changed more smoothly with brightness, showing good color stability. Further, by changing the ZnSe thickness to optimize the device, when the CPL was 45 nm, the maximum current efficiency and maximum brightness of the device were 7.38 cd/A and 1410 cd/m2, respectively, and the color coordinates were (0.30, 0.33).

To adapt to the development trend of high-spatial-resolution of uncooled infrared detectors, the uncooled infrared focal plane array (IR FPA), which is the core component of uncooled infrared detectors, is constantly developing towards larger arrays and smaller pixels. Aiming at the diode-type IR FPA, this paper theoretically analyzes the influence of the sensitive element diode on the readout circuit and device performance. While determining the best operating current of the diode, the number of series connected in the diode structure and junction area are the leading performance factors. Based on this conclusion, a p+n-pn-n+p 3-in-1 diode was designed and combined with traditional diodes, “well”-shape diodes, p+n-n+p 2-in-1 diodes, and two 3-in-1 diodes obtained by the direct expansion of p+n-n+p 2-in-1 diodes, which were compared and investigated. The study found that the p+n-pn-n+p 3-in-1 diodes of the six structures had the most significant number of diodes in series under the same size, and the relative junction area was the largest. Using Sentaurus TCAD simulation, it was verified that under the same overall size, the voltage temperature sensitivity (TCV) value of the p+n-pn-n+p 3-in-1 diodes is approximately that of the p+n-n+p 2-in-1 diode, and two types diode of 3-in-1 obtained by direct expansion on the p+n-n+p 2-in-1 diode was 1.5 times, which is 2.6 times and 3.7 times the “well” shape diode and traditional diode, respectively. It is proved that the performance of the p+n-pn-n+p 3-in-1 diode is the best under small pixels, and the expansion of the N-in-1 diode can further optimize the device performance.

A 1024×1024 EMCCD imaging sensor that captures images from sunlight to starlight was designed and fabricated. The pixel size was 10 .m×10 .m. The device includes nondestructive floating gate output amplifier, its charge-to-voltage conversion factor is 3.57 .V/e－, and its charge handling capacity is 55.04 ke－. The nondestructive floating gate output amplifier can sense the amount of charge present in a charge packet without altering the number of electrons in that charge packet. This enables the camera system to determine whether the pixel charge packet is routed through the normal gain output or the EMCCD output based on a user-selectable threshold. The intra-scene switchable gain feature enables wide-scale dynamic imaging. A vertical overflow drain structure suppressed the image blooming by 200×. This feature enables imaging under extremely low light, even when bright objects are within a dark scene, thus allowing a single camera to capture quality images from sunlight to starlight.

The effects of the microchannel plate (MCP) tilt angle on the MCP noise factor, resolution, and gain of an image intensifier was studied through an experimental comparison and analysis. The results showed that the relationship between the MCP noise factor and MCP tilt angle was parabolic when the MCP tilt angle was in the range of 5°–12°. The MCP noise factor was the smallest when the MCP tilt angle was 9°, and the resolution was negatively correlated with the MCP tilt angle. The resolution was the largest when the tilt angle of the MCP was 5°. The MCP gain changed parabolically with an increase in the MCP tilt angle. The MCP gain was the largest when the MCP tilt angle was 9°. Because the direction of the photoelectrons emitted by the photocathode of the image intensifier is certain, changing the MCP tilt angle causes the photoelectrons to enter the front end of the MCP channel at a different angle. This results in different secondary electron emission layer depths, such that the number of secondary electrons and the radius of the scattering spot formed by the electrons at the output end are different. To determine the best MCP tilt angle, the main performance requirements of an image intensifier under various usage scenarios must be comprehensively considered.

The main voltage heating equipment in UHV substations consists of 1000 kV CVTs, 1000 kV GIS outlet bushings, and 1000 kV arresters. They exhibit small differences in surface temperature and are easily exposed to environmental factors. Precise infrared temperature measurements are typically performed to diagnose defects under ideal environmental conditions. However, if the onsite environmental conditions cannot meet the requirements in a short time, further defects may occur. Therefore, it is necessary to study a correction method for temperature measurement results under non-ideal environmental conditions. In this study, three types of UHV heating equipment temperature field simulation models were established and combined with the theory of heat transfer to study the influence of environmental factors such as light intensity, temperature, and wind speed on the surface temperature difference of UHV heating equipment. The results revealed that the enhancement of these three factors decreased to varying degrees the surface temperatures that needed to be corrected to the true temperature difference under ideal environmental conditions. Finally, data fitting was used to obtain a curve expressing the measured and real temperature differences, and the correction method of the surface temperature difference of the UHV heating equipment under non-ideal environmental conditions. The accuracy of the correction results was verified through the field application of a substation.