Data processing apparatus for operating lens correction and method for compressing and restoring lookup table values

A data processing apparatus comprising:

• a plurality of register blocks, which are connected in series and are respectively capable of performing a data shifting operation;
• a shift block shifting first L-bit data, where L is a natural number, which is stored in a second register block among the plurality of register blocks, as K-bit data, where K is a natural number and less than L, and outputting shifted first L-bit data;
• an adder generating third L-bit data by adding the shifted first L-bit data and second L-bit data stored in a first register block among the plurality of register blocks and outputting generated third L-bit data to a last register block among the plurality of register blocks;
• and
• a switching block outputting a third L-bit data, which is shifted to the first register block, in response to a control signal output from a controller whenever the third L-bit data is shifted to the first register block through a data shifting operation.

A method for compressing and restoring one-dimensional data, the method comprising:

• setting an initial seed vector corresponding to an initial condition for a Cauchy problem that defines a polynomial regression to compress a plurality of original function values sampled from an original function in a lookup table;
• generating a seed vector corresponding to each of the original function values based on the set initial seed vector via a seed vector expansion operation;
• and
• restoring a corresponding original function value based on a corresponding seed vector of the seed vectors generated in the seed vector expansion operation.

A method for compressing and restoring two-dimensional data, the method comprising:

• setting a first seed matrix to compress function values in a two-dimensional matrix form sampled from a function having two arguments;
• generating a second seed matrix using a one-dimensional recursive near polynomial (RNP) expansion based on each of first arrays of the first seed matrix;
• and
• restoring the function values in the sampled two-dimensional matrix form using the one-dimension RNP expansion based on each of second arrays of the generated second seed matrix.