Research On The Theory And Design Method For Light Field Imaging

In the field of imaging, because of the lack of space information of optical axis position and the limit of a single focal plane, the traditional imaging technology has been unable to meet modern requirements in the high-dimensional area. Firstly, traditional imaging techniques can only record the projected information of the incident beam in three-dimensional scene; the information on the other dimension cannot be record. Secondly,traditional camera can only obtain an accurate focusing picture of a single object plane and it is easy to out of refocus; and we should to refocus mechanically when taking a picture. Multicast intersection measuring can be calculated depth information of the object, but because of calibration difficulties, camera settings and restrictions Pose intersection region makes it suitable only for special occasions, to affect the timeliness of the imaging system. In order to overcome the deficiencies of traditional imaging, the subject has learned the light field photography based on the microlens array.Firstly, we simulate system of the light field imaging by analyzing the basic theory of light field photography, which mainly comprise the parametric characterization of light field and so on. Secondly, through the spatial domain and frequency domain refocusing algorithm to deal with the light field data after a single exposure; and in order to find the precise refocusing image, we introduce an ideology that clarity evaluation function. we analyzes and compares the performance of these sharpness evaluation function by a series of digital focusing pictures. In light field camera design, the first we design the light field sensor through the selection of microlens array and detector and the coupling distance demonstration. Then we corrects the dark current and the microlens center of the acquisition image through using a relatively radiometric calibration method Finally, we designs a light field camera based on the simulation of light-field imaging systems. Experimental results of the digital refocusing process through the camera proved that the clarity of the evaluation function based on the frequency domain have a high sensitivity and a good anti-noise performance compared to the spatial domain, and achieve a better evaluation results. The results can also prove the accurately focused image corresponding to the maximum value of the evaluation function. So that through the evaluation function curve we can determine the precise focus of the image.