초록 ▼

The method identifies a pixon, a fundamental and indivisible unit of information, and a pixon basis, which represents the smallest number of cells of arbitrary shape or position required to fit the data, representing the minimum degrees of freedom necessary to specify the image. The goodness-of-fit

The method identifies a pixon, a fundamental and indivisible unit of information, and a pixon basis, which represents the smallest number of cells of arbitrary shape or position required to fit the data, representing the minimum degrees of freedom necessary to specify the image. The goodness-of-fit is determined according to the Maximum Residual Likelihood (MRL) statistic or a similar chi-squared statistic.

대표청구항 ▼

[ We claim:] [1.] A method for reconstructing a detected image from input data carried by a detected image signal generated by a detector array having a plurality of detector elements using Bayesian estimation, said input data comprising a plurality of data points, the method comprising the steps of

[ We claim:] [1.] A method for reconstructing a detected image from input data carried by a detected image signal generated by a detector array having a plurality of detector elements using Bayesian estimation, said input data comprising a plurality of data points, the method comprising the steps of:maximizing a goodness-of-fit between said input data and a most likely image given an estimated model;identifying an image prior by determining a most likely model given the most likely image;selecting a plurality of generalized image cells within the image prior, said plurality of generalized image cells corresponding substantially to portions of the detected image, each generalized image cell having a size, a shape and a position within said detected image based upon a local signal-to-noise ratio within said detected image signal;repeating the steps of maximizing a goodness-of-fit, identifying an image prior and selecting a plurality of generalized image cells until the quantity of said generalized image cells is the smallest quantity of said generalized image cells capable of containing the greatest possible collective quantity of said data points and a maximum goodness-of-fit threshold can be reached; andproviding an unblurred output image representative of said detected image with the blurring removed.