Microwave filter is the key building block in the radio frequency (RF) system such as wireless communication and radar. Recently, silicon integrated microwave photonic filters (MPF) have been a hot research topic due to their capacities on integration, large bandwidth, excellent tunability and immunity to electromagnetic interference. In this paper, the operation principle of MPF is introduced firstly. Then, the latest research progresses of incoherent and coherent silicon integrated MPF are reviewed, and also the problems and challenges are analyzed. Finally, the main achievements of silicon integrated MPF up to now are summarized, and the next research orientations are discussed.

Tunable external cavity diode laser(ECDL)has excellent characteristics such as narrow linewidth, wide range tuning, single-mode output and high output power. Tunable ECDL is widely used in white light interferometry, wavelength division multiplexing system, coherent optical communication, optical fiber sensor technology, etc. In this paper, the basic principles of tunable ECDL and the tuning mechanisms of various typical external cavity structures are introduced firstly. Then the latest domestic and overseas research progresses of various tunable ECDL are expounded according to different external cavity structures, and the advantages and disadvantages are also analyzed. Finally, the existing shortcomings and future development trends of tunable ECDL are summarized and prospected.

By using the gate control characteristics of GaN high electron mobility transistor (HEMT) and the photovoltaic effect mechanism of ferroelectrics, a new type of (photosensitive layer/HEMT) photo-detecting device was prepared. In this paper, mainly studied is the influence of film types and sputtering atmosphere on the photovoltaic performance of light-sensitive films and the light detection ability of the new photosensitive grid detector. The results show that the peak quantum efficiency of the PZT/ZnO composite film reaches 14.55%, and the residual polarization intensity of the PZT film prepared by passing oxygen during sputtering reaches 52.31μC/cm2. The saturation current of the detector deposited with PZT/ZnO composite film under UV light increased by 12.64mA at most compared with that of the dark field source. It can be concluded that the new detector has an excellent ability to detect ultraviolet light, and it provides a new direction for the study of light detection.

A 640×480 pixels interline transfer CCD with open window ability is designed and fabricated with the pixel size of 7.4μm×7.4μm. The device has four open window modes, including 640×480pixels, 228×480pixels, 640×164pixels and 228×164pixels. The device can output signals in single or dual output channels. When the window resolution of dual output images is 228×164pixels, the frame frequency can reach 500f/s, which meets the application requirements of high-speed target recognition.

In order to enhance the resolution of the flow sensor, a biomimetic hair flow sensor based on bi-stable potential energy adjustment is proposed. The sensor is mainly composed of hair, electrode layer, top structure layer, middle connecting layer and bottom supporting layer. On the one hand, the top structure layer adopts asymmetric comb fingers to realize differential capacitance detection; on the other hand, periodic modulated comb fingers are arranged to realize bi-stable potential energy function of the system. Through the fluid-solid mechanical-electrostatic multi-physical field coupling analysis, the relation of capacitance changing with the air velocity was obtained, and the resolution was better than 0.001m/s. In addition, the potential energy function curve of the system was drawn through the electrostatic field analysis of the modulated comb fingers. Finally, based on MEMS processing technology, a reasonable process was designed and the sensor was manufactured.

A method for fabricating optical fiber Fabry-Perot (FP) pressure sensor by using CO2 laser to melt capillary end face is proposed. The F-P structure is composed of ordinary single-mode fiber (SMF) and a silicon film prepared based on the end face of a CO2 laser fusion capillary. In this paper, the manufacturing process and melting parameters of the silicon thin film are introduced in detail and performance tests are performed on the F-P pressure sensor. The results show that the fabricated fiber F-P sensor realizes good sensitivity and linearity to the pressure, obtaining a sensitivity of 16.37pm/kPa, and the linearity of 0.998.

CsPbBr3 perovskite quantum dots were prepared by room temperature synthesis, and the surface ligand density was controlled by the cleaning times. By preparing the device structure as ITO/PEDOT∶PSS/PTAA/CsPbBr3 QD/TPBi/LiF/Al, the influence of cleaning times on the electroluminescent device was studied. The results show that when the quantum dots were cleaned twice, they can get balance between the charge injection and liquor stability, and a green perovskite QLED with a maximum brightness of 1405cd/m2, an external quantum efficiency of 0.6%, and a color coordinate of (0.127,0.559) was obtained.

In this paper, various surface structures of electron multiplier layer are designed, and the influence of doping distribution on the charge collection efficiency of electron bombardment CMOS (EBCMOS) imaging device is simulated and analyzed. Combined with ion implantation process, the surface structure of electron multiplier layer was optimized, and the influence of different kinds and thickness of masking layer, ion implantation dose, implantation angle and implantation times on doping distribution was analyzed by using ion implantation simulation software TRIM. According to the carrier transport theory and Monte Carlo simulation method, the charge collection efficiency of electron multiplier layer in EBCMOS is simulated and analyzed. The simulation results show that the charge collection efficiency can be improved by selecting SiO2 as the masking layer, reducing the thickness of the masking layer and increasing the number of energy injection. In the aspect of injection dose selection, heavily doping the surface layer of the electron multiplication layer can make the doping concentration decrease large enough and slow enough, which can also effectively improve the charge collection efficiency. After optimization, the charge collection efficiency of the device corresponding to the surface structure can reach 93.61%.

A kind of long linear array CMOS image sensor based on the capacitors trans-impedance amplifier (CTIA) is designed. A single-ended four transistors cascade operational amplifier is employed for eliminating the power dissipation and area. In order to promote the signal readout ratio, a minimized PMOS source follower without body effect is designed. The effect of the parasitic capacitance of the output bus is reduced. On the layout, the top metal is used to reduce parasitic resistance and capacitance. The signal readout ratio and the linearity range are raised. The image sensor chip with the pixel array of 5×1030 and the pixel size of 20μm×20μm is fabricated with 0.35μm standard CMOS process. Test results show that the image sensor works well when the integration time is 1ms and the readout ratio is 4MHz. The linearity can reach up to 98% and the linearity of dynamic range is 76dB.

Based on the optical and semiconductor modules of COMSOL software, the geometric design of spherical silicon solar cell was simulated from the three parts of diameter, spherical defects and anode area. By comparing the reverse saturation current density and the ideal factor, it is found that the smaller the radius of the spherical silicon, the smaller ball defect ratio, the larger the relative contact area of the positive electrode, and the better the electrical properties. The correlation between the geometrical characteristics and photoelectric parameters of the spherical silicon solar cells with different diameters was analyzed, and it was found that there were significant differences between the cells and the conventional planar silicon solar cells.

A 3D micro-electromechanical systems (MEMS) foldable structure for Micro-IMU application is proposed, which is composed of SOI substrate, polyimide flexible chain, metal wires and MEMS multi-ring gyroscope. In this paper, the feasibility of foldable structure and sensors is analyzed based on finite element method. Based on SOI integrated MEMS technology, the fabrication process of core sensor and foldable structure technology are integrated. The structure is equipped with a single-axis disk multi-ring MEMS gyroscope, and the flexible hinge is used to realize the 3D folding of the structure and the electrical interconnection between the sensors. A 3D foldable system with a volume of 1cm3 and a mass of 250mg was fabricated.

The evolution of finite array Airy beams (FAAB) propagating in gradient refractive index medium (GRIM) is investigated based on light transfer matrix and generalized Huygens-Fresnel integral equation. Firstly, the general propagation formulae of the Airy beam and FAAB passing through the paraxial optical system are calculated in space domain, respectively. Then the incident FAAB intensity profiles changed with exponential truncation and beam wrist size factors are discussed. The intensity characteristics and the periodicity of the transmission of the FAAB on several cross-sections of the GRIM are researched. The propagation behavioral differences between two kinds of beams are compared. The results are of theoretical significance to the transmission and communication of the array beams, because both the atmospheric random dither and graded index optical fibers can be abstracted as gradient index models.

The remainder particles in empty-sealed optocoupler will seriously affect its reliability. Particle impact noise detection (PIND) is the main method for detecting the remainder particles, but it presents an undetected error rate. Based on the voltage sampling data in PIND test, the detection probability model is established in this paper. The effects of distribution region and test frequency on the detection probability of typical kinds of particles with the same mass were studied by using the model. The results show that the detection probability of ceramic particles and conducting resin particles are greatly affected by hardness of impact object and particle adhesion, which makes the detection probability in different regions present obvious differences, and the overall detection probability of the latter is small due to the weak elasticity of particles, but will obviously improve after the test frequency rises. The results also show that the detection probability of short gold wire is large and stable due to its weak adhesion and strong elasticity. Based on this, the control and detection method for typical remainder particles is proposed, which will effectively improve the reliability level of the optocoupler. The model and research method in this paper can be extended to similar research on other empty sealed components.

Nano-TiO2 dispersion liquid was prepared in deionized water with ultrasonic methods by using nano-TiO2 as the raw material, and selecting sodium hexametaphosphate (SHMP), Tritor-X-100, polyethylene glycol (PEG) 400, sodium dodecyl sulfate (SDS), sodium dodecyl benzene sulfonate (SDBS) and cetyl trimethyl ammonium bromide (CTAB) as the dispersants. The absorbance of the sample was measured by UV-Vis spectrophotometer, and the influence of the ultrasonic time, the type and concentration of the dispersant on the stability of the sample was analyzed, and the optical performance of the sample was tested. The results of the study show that, stable nano-TiO2 dispersion liquid can be obtained by adding SHMP and ultrasonic treatment for 35 minutes with an absorbance of about 0.1, and under the illumination of the light source (CCT=10000K), the Rayleigh scattering effect of the simulated sky light can be achieved.

Aiming at the problem that the infrared image of rotary kiln surface is fuzzy and the image details are not easy to distinguish, an infrared thermal image enhancement method is proposed based on the visual characteristics of rotary kiln surface. In this method, the infrared image gray level of the rotary kiln surface is mapped to the easily distinguishable area of the human eye by utilizing the difference of the resolution ability of the human eye in different gray levels, so as to improve the resolution ability of the human eye to the image details. The experimental results show that the method can highlight the edge of infrared image and reflect the high and low temperature area in the kiln in time. It is of great significance to avoid the red kiln accident caused by abnormal high temperature and ensure the kiln to operate economically and efficiently.

An image matching algorithm based on KAZE and Perceptual Hash is proposed to solve the mismatching problem in image feature matching. Firstly, it uses KAZE algorithm to extract and match the image feature points. Then the perceptual Hash algorithm is used to filter and purify the initial matching pairs. Finally, random sampling consensus (RANSAC) algorithm is adopted to remove the false matching points, and thus the final matching results are obtained. Experiments performed on Mikolajczyk standard dataset indicate that the matching accuracy of the proposed algorithm is high with image changes in rotation, scale, illumination, blurring and JPEG compression, and has good reliability and robustness.

An adaptive model is proposed for gaze estimation based on differential convolutional neural network. The adaptive model incorporates information of head pose and designs a network named Differential Network (DNet) by virtue of differential convolution. The DNet is trained to predict gaze differences in the eyes, calibrate the initial gaze estimations and thus reduce the estimation errors. Through validation on the publicly available dataset Eyediap and comparison with other well-performed gaze estimation models developed in recent years, the experimental results indicate that the proposed adaptive model can estimate gaze directions more accurately under free head movement.

Aiming at the problem that traditional Census transform is susceptible to noise and the matching accuracy is low in the matching cost calculation process, a four-state Census transform algorithm with noise tolerance is proposed. In the calculation of the matching cost, firstly, the improved Census matching cost is fused with the matching cost of the absolute difference of grayscale and gradient, and the corresponding cut-off threshold is added to improve the reliability of the initial matching cost volume. Then, the cost aggregation is performed through the guided image filtering, and the strategy of “Winner Takes All” is used to calculate the initial disparity value. Finally, the final disparity map is obtained by optimizing the initial disparity value through left-right consistency checking, disparity filling and weighted median filtering. Experimental results show that the noise robustness of the proposed algorithm is better than that of the traditional Census transform algorithm, and the overall error matching rate of the stereo matching algorithm is reduced to 5.59%.

Aiming at the problems of high power consumption, high jitter and long locking-time of traditional phase-locked loops (PLL) in image sensors, a kind of fractional frequency all-digital phase locked loop (ADPLL) is proposed based on a counter architecture. It realizes low power consumption, low noise, low jitter and fast locking design. First of all, the dynamic adjustment lock control algorithm is adopted to reduce the loop noise and shorten the locking time. Secondly, a universal unit is designed to realize the integration of digital time converter (DTC) and time digital converter (TDC), reducing the jitter caused by the gain mismatch. Simulation results based on 180nm CMOS process show that the ADPLL can achieve frequency output in the range of 250MHz~2.8GHz under 1.8V power supply voltage, the locking time is 1.028μs. When the offset carrier frequency is 1MHz, the phase noise is -102.249dBc/Hz, and the root mean square jitter is 1.7ps.

Due to the advantages of safety and reliability, high accuracy and large range, laser thickness measurement is widely used in online measurement of film thickness such as paper and battery pole pieces. Scanning thickness measurement of the strip in the wide direction inevitably introduces mechanical vibration, which causes deviations in the measurement results. The lithium-ion battery pole piece was taken as an example to analyze and optimize this problem. First, the thickness of the battery pole piece and the copper foil was measured with differential thickness measurement system, then the spectrum of the battery pole piece and the copper foil was analyzed, and the similarity of the vibration laws was explored. Finally, based on the spectrum analysis, a filter was designed to process the thickness measurement data. After filtering, the range of the pole piece and copper foil are reduced by 33.4% and 73.8%, respectively, thus the thickness measurement accuracy can meet the requirements.

A method for re-amplifying and re-shaping optical signal based on four-wave mixing effect (FWM) in fiber is proposed to realize optical signal 2R regeneration, and the working principles of this method are analyzed. An optical fiber transmission simulation system based on FWM effect to realize 2R regeneration of optical signal is designed. Simulation results show that, the 2R regeneration signal is good and the system Q value is up to 21.59, which verifies the feasibility of the proposed method. The wavelength and power of pump light are the main factors affecting FWM efficiency and regeneration performance. The optimal performance of the system is obtained when the pump wavelength and power is 1548nm and 40dBm, respectively.

In this paper, the classification algorithm is used to classify and recognize ten types of targets and three types of target variants on the MSTAR data set. Then, according to the prior knowledge in the classification adjustment process, the evidence is corrected, namely the output of the classifier, and the basic confidence function is constructed. The improved combination rule is a conflict measurement method based on the combination of the conflict coefficient K and the Pignistic probability distance, and the conflict evidence is synthesized by the combination rule that distributes the conflict degree in proportion. Without fusion, the highest classification accuracy rate of the three types of target variants is 93.553%. After fusion, the classification and recognition rate of the three types of target variants is 95.092%, which is increased by about 37% of that of the ideal state.

A fast simplified SC decoding algorithm based on the uniform quantization is proposed to improve the problems of the high decoding complexity for the existing SC decoding algorithms in polar codes. The proposed algorithm retains the integer operation of the original SC decoding algorithm based on uniform quantization to save a lot of the storage space and facilitate the hardware implementation. Furthermore, the proposed algorithm reduces the operation time complexity by means of adding the identification of the special node. The simulation results show that the proposed fast simplified SC decoding algorithm, compared with the original algorithm, can reduce the time complexity by 46.29%. At the same time, the decoding performance of the proposed algorithm is only 0.1dB lower than that of the original algorithm at the block error rate of 10-5.

By introducing grayscale preference parameters and gradient valley coefficients, an improved grayscale preference valley emphasis (GPVE) method is proposed. The grayscale preference parameter makes the threshold show a tendency to approach the object gradation, and the gradient valley coefficient makes the threshold closer to the valley. GPVE method can achieve an ideal threshold when the histogram is unimodal and bimodal. The experimental results show that the average segmentation error ME value obtained by Otsu, VE, NVE, WOV and other methods is above 0.2, while the ME value of the proposed method is only 0.007. Thus the proposed GPVE method improves the accuracy of image segmentation and realizes a better segmentation effect.

To improve the efficiency and accuracy of automatic assembly of chip level atomic clock, a coarse-fine combined micro device pose detection method is proposed. Firstly, after analyzing the characteristics and application environment of the micro-devices, a rough localization method of micro devices based on SVM classifier trained by image gradient histogram features is proposed. At the same time, occlusion samples and fuzzy samples are constructed to solve the problem of missing detection in the case of defocusing and occlusion of micro devices, so as to improve the generalization ability of the model. Then, the LSD algorithm and linear growth method are used to extract the edge of the target and calculate the sub-pixel feature point coordinates of the image. Finally, the real-time position and posture of the micro device under the microscope field of view are calculated by the ICP algorithm. Experimental results show that the recognition accuracy of the core components of the device is 95.3%, the average positioning error is less than 26μm, the angle error is less than 0.5°, which can meet the needs of the robot to accurately grasp the targets in the micro assembly process.

In the LiDAR system, due to the variation of the target distance and reflection characteristics, the peak value of the echo signal fluctuates in a relatively large range, which leads to a big fluctuation of the arrival time of the echo determined by the constant ratio timing discriminator circuit, and the resulting walking error affects the ranging accuracy and measurement range of the system. In this paper, a constant fraction discriminator with adaptive gain control is designed, thus the gain of the amplifier circuit can be adjusted quickly and automatically based on the level of the echo signal, which can adapt to the target measurement in a large dynamic range. The results illustrate that the designed circuit adopting automatic gain and constant fraction discriminator component is able to automatically configure and amplify the circuit gain within 25ns. When increasing duration of input signal to 3ns and the voltage dynamic range is stabilized around 29dB, the automatic control system adopted automatic gain design can effectively reduce the fluctuation range of dynamic error and the discrimination error of echo signal to 9dB and 0.42ns respectively, reducing the walking error significantly.