Study And Application Of Error-compensating Algorithm In Phase-shifting Interferometry

The technology of phase-shift interference is researched as a kind of measuring in interference computation. In this field, several new phase- shifting methods in practicing and computing are advanced. Because its precision is high, and it can define any point in the square accurately. That is to say, the method can practice the whole field orientation. And its performance contains several other characteristics, such as little influenced by the quality of the interference pictures, never limited by the shape of the stripe, high computing speed and realizing the real-time measuring. Because of the strong-points depicted upwards, the method becomes one of the main interference computation and it is widely used in several fields, such as the measurement of 3-D shape, the technology of digital micro holography, the detection of coarseness, the encryption of images and the technology of water marked.The precision of phase-shifting in this measuring technique or not has the very tremendous influence to the measurement result accuracy. Although many algorithms have been appeared to be able to suppress the phase-shifting error, these methods all more all less have the flaw. So to eliminate the phase-shifting error well which in the phase-shifting process have the certain difficulty.This article introduces some commonly used phase extraction algorithms, carries on the comparison to each algorithm, and presents one algorithm with the cosine fitting to the phase distribution. It make the comparison and computation in the same region interference fringe to solve the actual phase value and the phase-shifting quantity, then produces the correction computation formula through the relation between the phase construction and the phase-shifting errors, and achieved the suppression phase-shifting error's goal. The article presents the new algorithm formula and describe phase-shifting error solution process, then give the computer simulation result and the experimental survey data processing, at last, confirmed the new algorithm is validity.