The solid-electrolyte based metal-oxide thin film transistors has huge potential due to its good environment stability and excellent electrical performance. However, the sophisticated fabrication process and time consuming are the main drawbacks involved in the traditional method to modulate the electrical performance of solid-electrolyte based metal-oxide thin film transistors. Laser processing on the materials provides excellent processing speed and precise control. The metal-oxide thin film transistor was realized utilizing high k solid electrolyte as dielectric layer and transparent indium tin oxide (ITO) as semiconductor layer. Prior to the deposition of active layer and electrodes, fs-laser was induced to irradiate on the surface of Ta2O5 dielectric layer. As the intensity of laser irradiation increased, the ION of the transistor was enhanced and the threshold voltage was negatively drifted. In addition, the Excitatory Post Synaptic Current(EPSC) of the solid-electrolyte based metal-oxide thin film transistors was studied under different laser intensities. The EPSC was significantly increased as the laser intensity enhanced. Meanwhile, XPS results showed that the amount of oxygen vacancies could be controlled via laser irradiation. A simple, fast(<1 s) and low-temperature(<45 ℃) method was proposed to modulate the performance of thin film transistors.

Voice coil actuator(VCA) has the merits of no hysteresis, fast response and large stroke. It has been used in the secondary mirror of large optical telescopes. To improve its efficiency, a high efficiency moving magnetic voice coil actuator was designed. And a three-dimensional model of VCA has been built and analyzed with finite element analysis software to simulate its magnet magnetizing direction, magnet size, coil size and structure contributed to a more accurate result of the electromagnetic force and efficiency. In order to generate large electromagnetic force and high efficiency, the above parameters are simulated and optimized respectively. The result shows that the output force of the small caliber moving magnetic VCA is up to 0.43 N and the actuator constant is up to 0.9. There is a good linear relationship between the output force and the input current, which provides a theoretical basis for the development of the VCA for next generation deformable mirror.

Small interval LED based naked eye 3D display system is complementary to LCD based naked eye 3D display system for its relative high resolution and large display size. The sub-pixel of LED is point-like and can be arranged in horizontal or vertical directions. In this paper the effect of sub-pixel direction on cross talk of display system is investigated. The Tracepro software is used to analyze crosstalks of two 3D display systems with sub-pixel interval as 1.25 mm and 1.667 mm, respectively. The simulation results shows 3D display systems of different sub-pixel arrangement directions have nearly the same crosstalk. This result is further confirmed by experimentally testing crosstalk of 1.25 mm LED naked eye 3D display system. With 1.25 mm LED naked eye 3D display system, its sub-pixel arrangement direction can be changed freely according to actual requirements.

Transmission is an important optical parameter in TFT-LCD products. In order to reach some optical requirements, such as view angle, the transmission axis on polarizer film of TFT-LCD product is changed. However, there are obvious differences in the transmission of the TFT-LCD products which have the same bare cell design but different transmission axis on polarizer films. The polarizer film and the backlight unit should be matched. In this paper, the experimental and theoretical studies show that the polarization property of backlight unit is the primary factor on the obvious differences in the transmission. The polarization property is mostly influenced by the prism film. The main polarization direction is approximately perpendicular to the prism direction. In the condition of 3° prism backlight unit structure, the transmission difference and the degree of polarization of backlight unit are linear relations. The linearity is more than 0.99.

The undesirable phenomenon of edge dark band of TFT-LCD is introduced, the mechanism and influence factors are studied, and the optimum process parameters are concluded. The experimental results show that, from the perspective of panel design, the uniform thickness of edge cell gap is helpful to improve the uniformity of brightness in the edge area, and therefore the moderate amount of liquid crystal, the moderate size and doping ratio of silicon ball, and harder sealant are preferred. Meanwhile, the flatter the panel after polarizer attachment, the lighter the degree of dark band. From the perspective of backlight unit, the edge brightness uniformity of finished products has strong correlation with the batch of backlight. However, this is mainly related to the dimension parameters such as flatness of the backlight rather than the degree of dark band of the backlight. In addition, the aging process releases the internal stress generated during the assembly of backlight with panel, and alleviates the deformation at panel edge. Therefore extending the aging time appropriately is conducive to reduce the incidence of edge dark band. Through the application of optimum conditions, we have successfully improved the uniformity of edge brightness of our product and effectively enhanced its competitiveness.

Advanced super dimension switch(ADS) technology have many advantages, such as high resolution, high transmittance, low power consumption, wide view angle, high aperture ratio, low color shift, no push Mura, etc. Compared to VA(Vertical Aligned) mode, low contrast ratio is the disadvantage of the ADS mode. New pixel design & new LC(Liquid Crystal) for CR(Contrast Ratio) improve ment, is proposed in this paper. By using high transmittance pixel design, L255 luminance is improved and L0 has no change， the CR is improved by 8.3%(Test on 55UHD & 49UHD Product). By developing low scattering LC, L0 luminance is decreased and L255 has no change, the CR is improved by 22%(Test on 49UHD Product). CR is totally improved by 30%.

The traditional LED intelligent glass display control system needs a computer as the video source for real-time video output, but it is not convenient to use in small-scale occasions, so this paper designs a portable video playing control system which is separated from the real-time video source. The system is mainly composed of two parts: video output unit and video forwarding unit. Video output unit adopts CPU + FPGA dual core architecture based on operating system. The CPU reads the video data in the USB disk through the USB interface, draws the address mapping map in the specific built-in APP of CPU, generates the address mapping data, and sends the video data and mapping data to FPGA. FPGA reorders the video data and sends it to the video forwarding unit through optical fiber. The video forwarding unit receives and selects part of the video data. After anti Gamma correction, it is sent to the driving system to realize the control of LED on the display screen. Multiple video forwarding units are used to drive large-scale LED intelligent glass display screen in cascade. This system brings convenience to product application and has a wider range of application.

The principles and developments of display driver chips in recent years are summarized, and their development trend in near future is prospected. Firstly, we present several common display screens first, introduce the principle of display driver according to the type and characteristic of different display screens, and list the current development status of display driver chips. Then we sort out the development highlights of recent years and development trend. And lastly, we forecast the future development of the display driver chips. By comparing the display driver chips of various popular companies in recent years, we found that the innovation speed of display driver chips has increased significantly and the functions have increased extensively. As the fast changing of display technology, display driver chips will also develop more rapidly and change dramatically in future years.

Due to the large number of model parameters, common target detection models were often difficult to be deployed on mobile embedded platforms such as unmanned aerial vehicle and satellite. In order to detect ships in real time, and for the purpose of deploying target detection model in weak computing equipment, the ship detection algorithm based on computer vision was researched. According to the feature of ship shape length ratio and width ratio in satellite images, K-means ++ clustering algorithm was used to select the initial candidate anchor boxes. Multi-scale pyramid images were used as the input of model training. The scale factor of the batch normalization layer of the YOLO-v3 target detection algorithm was taken as the measure index of channel importance, and the YOLO-v3 model was pruned and compressed. Experimental results show that model pruning and compression method can effectively compress the model. The number of parameters of the model size is reduced by 91.5% and the time of model detection is shortened by 60% compared with the original model, which greatly reduces the overhead of system computing performance. When the initial number of candidate boxes is 6, the mAP reaches at 77.31%, which meets the requirements of real-time detecting ship in satellite images.

At present, the research of camera imaging simulation mostly has the problems of less imaging models and incomplete coverage, which is difficult to meet the simulation requirements of camera in complex observation conditions. Therefore, an imaging simulation scheme of space optical camera based on the combination of satellite tool kit and open scene graph was proposed. Satellite Tool Kit(STK) was used to establish the orbit parameters of the observed object and the camera, and through data import, 2D and 3D scene information of the observed object surface is created. Through the analysis and research of many influence models, the influence model is created in Open Scene Graph(OSG) to realize the camera imaging process simulation under complex observation conditions. The distributed technology was used to design the system structure and complete the development of the whole simulation system in modules. The experimental results show that the system can provide a good simulation environment for the imaging process of space optical camera under the influence of complex conditions, and provide powerful data support for the design and optimization of space optical camera.

In order to achieve the purpose of real-time imaging processing of synthetic aperture radar (SAR) data, based on the use of spatial light modulator (SLM), a SAR data real-time imaging oblique plane optical processor is designed, and the optical path is adjusted to ensure image quality and reduce the volume occupied by the system. First, SLM is used to replace the film in the traditional oblique plane processor to reduce the photoelectric conversion rate of SAR data. Then, the structural parameters of the optical processor are set according to the known SAR related data. According to the horizontal and vertical reduction ratio of SAR, a cylindrical lens that meet the required magnification is designed, and a spherical lens is designed according to the overall length of the cylindrical lens to ensure that the cylindrical lens has sufficient working space. After the design is completed, the optical path is adjusted to reduce the volume occupied by the system, which is compact and lightweight for easy use on spaceborne or on board. The design results show that the system MTF is greater than 0.4 in the cut-off frequency, which meets the imaging requirements. After the optical path is optimized and adjusted, the total length is changed from 1 400 mm to about 700 mm. The results meet the design requirements and actual use requirements.

In order to solve the problems of low identification efficiency and poor real-time performance to cotton foreign fiber detection, an improved method of cotton foreign fiber detection based on YOLOv3 was proposed. The constructed model of MobileNets-YOLOv3 was designed, in which a lightweight MobileNets network was selected as the feature extraction network, combined with YOLOv3s multi-scale feature fusion detection network. A segmented learning rate was proposed to enhance the learning effect. The cotton foreign fiber image data set collected from the equipment on real site was divided into a training set and a test set, according to a ratio of 4∶1. Six image augmentation methods, such as contrast enhancement, brightness enhancement and so on, was induced, to expand the original data set. Based on the training set before and after expansion, different learning rates, the YOLOv3 model before and after improvement, the proposed model and Faster R-CNN along with SSD~~300, different models were tested and compared. The experimental results showed that the data expansion and the segmented learning rate can improve the overfitting of the training model, and the mean average precison (mAP) on the test set was increased by 3.6% and 5.64% respectively. The improved YOLOv3 model can significantly reduce the mistake rate and miss rate of detection. The mAP of the test set was 84.82%, and the frame rate was 66.67 f·s-1. Compared with YOLOv3, the mAP of MobileNets-YOLOv3 was increased by 2.03%, the frame rate was raised to 3 times as it and the duration of training was shortened to 1/4 times as it. The overall performance of MobileNets-YOLOv3 was also better than that of Faster R-CNN and SSD~~300. The proposed method can better satisfy the accuracy and real-time performance of cotton foreign fiber detection.

In order to solve the problem of monitoring and analyzing massive traffic video data, the in-depth research on traffic video surveillance technology in the context of hadoop big data is conducted, and a design scheme of anomaly jam detection algorithm is proposed based on traffic video data to realize traffic real-time data update and anomaly analysis. At the same time, for the massive traffic monitoring video, a parallel implementation algorithm is designed based on Hadoop component MapReduce. Finally, the effectiveness and accuracy of the algorithm is verified by actual traffic data of a city in Zhejiang Province. The algorithm in this paper can effectively calculate the traffic congestion and abnormal conditions. Compared with the traditional scheme, this scheme can focus on the time granularity in the range of 10 min to analyze the traffic situation in real time. Compared with the traditional distributed computing model, the 10 minute delay of this scheme can be controlled at 2.1 s, which is 81% lower than the traditional scheme, which basically meets the real-time, fine-grained requirements for traffic video surveillance requirements.