| Focusing system is an indispensable part of aviation visible zoom lens, especially in the flight process, due to the changes in temperature and barometric pressure, the vibration, altitude changes and other factors which cause defocus frequently, so image quality of zoom lens has a very close relationship with focusing system working normally and efficiently or not. Focusing system consists of three subsystems: the first part is an optical imaging system, which is used to capture the current image; the second part is the focus evaluation system, which judges the acquired images by some means(such as image processing method, auto-collimation method, etc.) to judge the defocus degree of system; the third part is the execution motor control system, which converts the defocus acquired by the focus evaluation system to the displacement of motor and completes the focusing by controlling the execution motor precisely so that the corresponding lens group can reach to a suitable position.Camera focusing system uses a cam structure to address the corresponding positional relationship between the focusing lens group and compensation lens group currently. However, relying simply on a cam structure can not meet the requirements of a clear image due to the special running environmental factors(temperature, pressure, altitude, etc.) of aerial camera, and so, other methods need to intervene to help focus to achieve clear images. Currently, aerial camera focusing method can be divided into manual focus and auto-focus, manual focus which needs people evaluate focusing according to visual feedback and long time to focus is inconvenience to capture moving targets, so the trends of focusing methods of aerial camera evolve to auto-focus gradually. One of the popular auto-focus methods currently is depth from focus, which is based on the process of a search to find the best imaging position, but there are some problems such as multimodal of evaluation curves in practical applications and the focus of the real-time requirements are not satisfied. In order to improve aerial camera auto-focus methods and the efficiency of auto-focus system, a novel method for autofocus is proposed in this paper, which shortens the focusing time compared to traditional methods, removes search process, and completes once orientation for focusing lens group. The method builds comparing templates by classifying the large number of defocused images acquired based on the analysis of 1-D power spectrum of image, and achieves the ideal filter function which can extracts the information of defocus degree of images through the analysis of overall experimental samples. The current defocus degree is obtained through the comparison between the result of the achieved filter function analysis on the image obtained currently and the templates established. Finally, convert the estimation of defocus degree to the displacement of the executive structure to complete focusing by driving focus motor to move the corresponding displacement. The results showed: the calculation time of defocus is 0.4s under the current experimental conditions, the proposed method has greatly improved the performance compared to 4 ~ 8s of depth from focus, which is a better real-time auto-focus method. Taking into account the effects of noise and illumination for focusing, in order to complete focus by the auto-focus algorithm successfully, denoising and low illumination enhancement will be performed before the defocus evaluation of images captured in this paper.Lens group focus, plane mirror focus and the image plane moving focus are the three most widely used in current focus systems. Lens group focus achieve focusing by moving the corresponding lens group; plane mirror focus approach adjusts the location of the plane mirror group of the system, and is applied in reflector system frequently; the image plane moving focus approach achieves focusing by adjusting the position of the image plane, and applied for image plane smaller systems. Lens group focus method is selected to adjust the internal lens group of original optical system in this paper, which reduce the overall system weight, improve the utilization efficiency of space equipments, and constitute a complete auto-focus system with the focusing algorithms based on image processing and electronic cam in this paper.In terms of executing structure, replace the mechanical structure of cam by dual linear motor in the traditional optical imaging system auto-focusing structure in this paper, and associate with robust control algorithms, so that the electronic cam structure ensure that the original mechanical cam accuracy, while reducing the size and weight of the overall structure and theoretically focusing lens group can be omitted. Due to poor aerial camera work environment and the varying degrees of interference in flight, a control system is designed by a servo motor with strong robustness control algorithm in this paper to solve the interference of motor controlling caused by environment issues. In this paper, permanent magnet synchronous motor was applied as motor execution of electronic cam, simulating experiment showed that the system do not overshoot, when the motor from rest tracking 600r/min speed setting, 2r/ min of steady-state accuracy; when the motor speed was 600r/min, adding 1.6N ? m of torque disturbance, maximum fluctuation of speed 5r/ min. Simulation and experimental results demonstrated the effectiveness of the control scheme about antidisturbance, robustness and steady-state performance, focus of auto-focus camera systems could be implemented through a fast and stable positioning of the corresponding lens group in electronic cam structure, good anti-interference ability. It could respond to the impact caused by environmental changes during the flight and other factors about motor control.The key issues of aviation zoom lens automatic focusing system need to be resolved mainly on two aspects: the one is defocus caused by the dynamic changes in planes flight environmental factors(temperature, pressure, distance, etc.). If the system adopts the traditional way based on sensor focusing, the indicators of sensors are demanded highly, moreover, due to the addition of the sensor makes the system design complexity, take up more space, and does not conform with aviation equipment miniaturization and lighter direction; on the other hand, the aircraft flying at high speed while the camera work, the scene changes quickly, which require the automatic focusing system with good real-time, and focusing work should be completed quickly and accurately. Image processing methods are adopted in this paper, preprocessing the images captured by lens at the first, namely denoising and low illumination enhancement; then,a study about the two terms referred above is launched, the system parameters disturbances caused by environmental factors change is solved by the research of focusing motor algorithm, a fast and accurate auto-focusing algorithm based on image processing improves the real-time of focusing system, no sensors and other devices added, and the dual linear motors replaced the mechanical cam structure, which reduced the overall volume and quality of the system, and meet the miniaturization and lighter requirements of aviation equipment. |
PDF Full Text Request