Due to the limited data transmission bandwidth between the Earth and Mars， it is impossible to transmit a large number of high-resolution images back to the ground in real time. The Tianwen-1 high-resolution camera requires the payload itself to have image storage capabilities. Aiming at the huge problem of high-resolution imaging camera imaging data， this paper designs a FPGA based NAND Flash image storage and processing system. First， according to the output requirements of CCD and CMOS image sensors， NAND Flash storage space is divided. Then， in order to solve the problem of high-rate storage， a pipeline storage method was designed. Next， aiming at the problem of newly added bad blocks on orbit， a self-checking method for bad blocks is designed. Finally， in order to solve the problem of tight transmission bandwidth between Earth and Mars， image processing methods such as down-sampling， pixel binning， and region extraction are designed. The test results show that the high-resolution imaging camera storage and processing system receives and stores the image data rate up to 3 Gb/s， and the download data rate can be reduced to about 4% of the original data rate. By using a variety of download working modes， it is possible to obtain rich image data and meet the bottleneck of deep space exploration data transmission.

In order to solve the problems of multi-state impact on camera performance and to maximize efficiency， multi-states and their transition models were investigated. First， the origin of multi-state was identified， and the effect of deterioration of each function unit was analyzed. Then， a model containing multi-state transition was established. Finally， a simulation method is put forward. A brief case study indicated that considering multi-states would improve the accuracy and reliability of prediction， and that sensitivity analysis will help in balancing the system reliability requirement and system reliability. The multi-state reliability simulation method can be a more accurate method for predicting the reliability of space cameras and provides a practicable engineering solution.

To obtain an image of the Martian surface with a high signal-to-noise ratio （SNR） and suitable brightness， the exposure parameters of the CMOS detector in the Mars high-resolution camera were designed， and the rolling shutter distortion was analyzed and corrected quantitatively. First， the photoelectric response model of the CMOS detector was established. By comparing the influence of exposure time and PGA gain on SNR， the exposure time was determined as the priority adjustment parameter. After analyzing the constraints of exposure time and clarifying the calculation method of target irradiance， the exposure time design results and SNR under different reflectivities， solar zenith angle， and orbit height parameters were given. Finally， the distortion mechanism of the rolling shutter was analyzed， and the correction method and theoretical correction accuracy were presented. The simulation results indicate that the SNR remains at approximately 48.28 dB in the non-orbital image-motion restricted area， and gradually decreases in the restricted area， with a minimum value of 32.45 dB when the reflectivity is 0.1. The SNR is 45.56 dB in a typical irradiation environment. Under the orbital parameters， the rolling shutter of the CMOS detector causes tens of pixels of tilt distortion. Under the premise that the image motion measurement error does not exceed 3.46‰， the image displacement after correction using the proposed method is better than 1 pixel.

In order to realize high-quality imaging of the Mars surface by the Tianwen-1 high-resolution camera in large elliptical orbit， a large elliptical orbit image motion calculation model based on Mars is established. The calculation method of the large elliptical orbit and image motion is studied. First， according to the characteristics of the large elliptical orbit， the difference between the large elliptical orbit and a near circular orbit is analyzed. Then， the distance from the high-resolution camera to the center of Mars， the orbital velocity of elliptical motion， and the angle between orbital velocity and centrifugal velocity are calculated according to the characteristics of the ellipse and the conservation principle of angular momentum. Then， according to the method of homogeneous coordinate transformation， the Mars scenic spot coordinate system is connected with the high-resolution camera coordinate system， and the image motion velocity is obtained by deriving the position relationship. Finally， the correctness of the image motion compensation calculation method and its influence on image quality are verified through a dynamic imaging experiment. The experimental results show that： the calculation method of image motion compensation for large elliptical orbit is correct； the dynamic transfer function is greater than 0.1， and the dynamic transfer function is not less than 0.1； and the minimum value of image motion matching MTF （Modulation transfer function） is 0.953， which meets the requirement of no less than 0.95.

To solve the problem of low bandwidth between Mars and Earth and the high-output image data rate of the high-resolution camera of Tianwen-1， an electronic image-storage system was designed by using a NAND Flash as a storage medium and a FPGA as a control core. The NAND Flash was designed to use parallel processing and 4-level pipelining technology to meet the requirements of high data rate storage of camera images. To ensure the reliability of the image storage system， the storage system area was divided and a bad block replacement mechanism was established. Moreover， to avoid the influence of the original bit error rate and the single event radiation effect of space， a 12×4 bit Hamming verification was proposed. The experimental results showed that the designed electronic image-storage system could achieve the highest real-time image storage data rate of 1 276 Mbps on each single channel， reliable storage of image data， and 1 bit error correction for a 12×4 bit unit of image data， meeting the requirements of the space mission.

In the context of deep-space exploration， such as the Tianwen-1 mission， reliability over a long working time is an essential requirement for the main control unit of the high-resolution camera. To improve the anti-single event effect and reliability of the master FPGA， an anti-fuse FPGA is used for the software system design. Functions such as focusing unit control， time code punctuality， and time correction， managed by the master FPGA， are safety-critical items. To improve the robustness of the software， the main control FPGA design employs a two-division control method based on the acceleration and deceleration curve for the optical-mechanical structure of the focusing mechanism. A stepper motor driver is used to match the focusing speed and avoid the resonance frequency of the mechanism， for the focusing mechanism to run smoothly； at the same time， when the Tianwen-1 surround is in the ring fire section， the high-resolution camera needs to perform shooting tasks by following delayed commands， creating a requirement for highly accurate timing. Thus， the master control FPGA design must ensure time code punctuality and timing function such that the high-resolution camera accurately executes the shooting task. The experimental results show that the movement time of the focusing mechanism is 112.2 ms and the accuracy of time keeping is 1.25 s. The design satisfies the requirements of a stable operation of the focusing unit and reliable timing.

To obtain high-quality images during the operation of the Tianwen-1 payload high-resolution camera， the imaging parameter settings were investigated. Combined with the results of laboratory calibration experiments， the default parameters of the camera in orbit and the imaging parameters under different ground reflectivities and sun altitudes were determined. First， the radiance of each spectral segment at the entrance pupil of the camera was calculated based on the atmospheric radiation transfer model. Then， the output charge value of the CCD detector was obtained by applying the photoelectric conversion model. Furthermore， the imaging parameters of the camera were obtained， and were found to meet the requirement of SNR > 100. Finally， spectral calibration and radiometric calibration were conducted under laboratory conditions， and the calibration error was analyzed. The experimental results indicate that the video response of the high-resolution camera is linear， relationship between the gray value and imaging parameter is linear， correlation coefficients of linear fitting are all above 0.999， non-uniformity of each spectral segment is less than 1%， and SNR is no less than 100 under typical lighting conditions. The relative radiometric calibration uncertainty is better than 3%， and the absolute radiometric calibration uncertainty is better than 7%. The results of the theoretical calculation and calibration are found to be basically consistent. Thus， the imaging parameters are designed reasonably， and the calibration results meet the system requirements.

The Mars orbiter orbits around Mars in a large elliptical orbit. Its high-resolution camera acts as one of the payloads of the Mars orbiter， and captures images of Mars from its orbit. The imaging quality of the camera is not only related to its own parameters， but also to the attitude and orbit control error. To realize high-resolution photography of Mars， the relationship between attitude and orbit control error and camera imaging quality is investigated in this study. First， the influence model of attitude and orbit control of the platform and image motion calculation of the high-resolution camera was analyzed. Then， the platform for the experiment and verification was built using a collimator and dynamic target simulator. The parameters of the orbit and satellite platform were simulated using ground simulation software， and the parameters were injected into the dynamic target simulator and high-resolution camera. The dynamic target simulator can generate the dynamic target， which is used as the shooting target by the camera， and the velocity of the target varies with the attitude and orbit parameters. The camera performs image motion calculation on real-time and imaged according to the image motion compensation parameters. The experimental results indicate that the dynamic MTF of the camera can be reduced by less than 10% when the orbit accuracy is less than 1 km and velocity accuracy is less than 1 m/s. High-precision attitude and orbit control can guarantee the imaging quality of the high-resolution camera.

The imaging quality of a space-borne camera is key to accurately obtain remote-sensing data. Exposure is an important parameter in the design of a space-borne camera， which affects the image quality of the camera. Auto-exposure obtains information from the target scene in advance and uses it as the basis for determining the exposure， which can avoid the loss of image information by overexposure or underexposure. In order to obtain high-quality image data from the high-resolution camera in Tianwen-1 under the complex illuminance of Mars， an auto-exposure imaging circuit， based on TDI CCD push-broom and CMOS staring imaging， and an FPGA-based auto-exposure algorithm is proposed. The auto-exposure experimental results from the high-resolution camera in Tianwen-1 show that the dynamic transfer function MTF of the image after auto-exposure is increased by 0.013. In an outdoor test， a comparison of the actual scene before and after the auto-exposure shows that the effect of auto-exposure is good， with better scene adaptation. The proposed circuit has the ability to meet the requirements of auto-exposure imaging under the complex illumination of Mars.

The Tianwen-1 high-resolution camera adopts an off-axis three-mirror optical system， with a high precision index， complex structure， and short development cycle. In view of the above requirements， the computer-aided alignment technology of the off-axis three-mirror optical system is thoroughly investigated and applied to the alignment process of the main optical-mechanical structure of the Tianwen-1 high-resolution camera. The application of this technology causes the alignment index of the main optical-mechanical structure of the camera to converge rapidly. The average system RMS is better than λ/14 for each field of view and the average transfer function is 0.381 at the characteristic frequency of 57.1 lp/mm. A series of environmental simulation tests， including a mechanical test， temperature cycling test， and weightlessness adaptability test， was conducted on the main optical-mechanical structure. Following the experiments， all the technical indexes changed only slightly， thus meeting the requirements of the overall design indexes. Finally， Tianwen-1 was successfully launched， and clear images of Mars were sent， thus proving the effectiveness of the computer-aided alignment technology.

The entry-pupil radiance of the optical imaging instruments of Tianwen-1 varies considerably during its orbit operation， owing to the changes in the solar elevation angle and ground scene reflectance. To achieve the best imaging effect， the optical imaging instruments should utilize on-orbit adaptive adjustment gain. The solar elevation angle is an important parameter that is used to set the integral series of the time delay and integration （TDI） charge-coupled devices （CCDs）. Furthermore， the integral series is the main parameter used to adjust the gain. This study presents a method for calculating the solar elevation angle in real time at the sub-satellite point of the Mars orbiter based on Fourier fitting， to mitigate the challenges of the real-time variation in the solar elevation angle and the large ephemeris file created during the orbiting period. First， an 8-order Fourier approximation based on the principle of least squares is utilized to fit the x， y， and z coordinates of the sun vector in the Martian inertial coordinate system， and a fitting equation is obtained as a function of time. Second， the real-time coordinates of the orbiter in the Mars inertial coordinate system are obtained based on the orbit parameters sent by the guidance and navigation control system. Finally， the solar elevation angle of the sub-satellite point can be calculated in real time on the orbit based on the cosine formula of the included angle. The experimental results show that the maximum absolute error of the real-time calculation results of the solar elevation angle obtained by using this method is less than 0.3° during the period from 2021-01-01 00：00：00 UTC to 2024-01-01 00：00：00 UTC. The accuracy requirements of the calculation results of the solar elevation angle of the TDI CCD integral series of the Tianwen-1 high-resolution camera are satisfied. Based on this method， the Mars image obtained by the high-resolution imaging camera of Tianwen-1 shows rich details with reasonable brightness and contrast.

In order to fully utilize the limited space and field of view in deep space exploration， a camera with three linear TDI CCD detectors is proposed. To acquire remote sensing images with high SNR and high-resolution， according to the chain of the imaging system， methods for analyzing the key factors that affect the SNR of high-resolution cameras as well as improvement measures are put forward. First， each channel is designed independently to reduce circuit coupling between channels. Then， the time-constant of the vertical transfer frequency is adjusted to reduce the crosstalk of the inner TDI CCD. Finally， the three detectors are physically isolated to reduce the near field coupling noise. The results of the radiation calibration experiment with the high-resolution camera in Tianwen-1 indicate that by simulating the orbital conditions of Mars， the SNR of the panchromatic spectrum is 115.1 when the integral number is 32， as the elevation angle of the sun is 30° and the surface albedo is 0.2. The results show that the design satisfies the criteria for Mars exploration.

To improve the mechanical and thermal performance of the onboard interface of Tianwen-1’s high-resolution camera， the structural design is determined by using inputs of the fundamental frequency of the camera and the satellite temperature. First， the onboard interface support requirements of a Mars high-resolution camera are analyzed， and common support methods for the onboard interface of an optical remote sensor are introduced. Second， the relationship between the optical performance of the remote sensor and the onboard interface is established. The sensitivity matrix of the spatial position change of the interface to the mirror surface shape is determined by performing a Monte Carlo analysis， and the interface design parameters affecting the stiffness and thermal stability of the entire machine are identified. Then， analysis models of the different interface schemes are established， and the optimal interface design scheme is determined. Finally， the fundamental frequency and stability of the entire machine under the selected onboard interface scheme are determined by conducting experimental tests. The test results show that with this interface scheme， the first-order fundamental frequency of the entire machine reaches 58 Hz， which is considerably higher than that of the satellite， and the MTF before and after the vibration are 0.196 and 0.187， respectively. Furthermore， when the temperature change in the onboard interface is 20 ℃， the MTF before and after camera focusing are 0.173 and 0.223， respectively. This spaceborne interface analysis and design method， based on statistical analysis， can effectively determine the interface design input， and the kinematics support solves the design problem of the interface of Tianwen-1’s astronomical camera. The design ideas and test results have considerable potential in guiding the design of this type of load interface.