Digital image processor

A digital image processor for processing input image data based upon a cellular neural network, comprising:

• first multiply-and accumulate arithmetic means for digitally processing multiplication and accumulation of input image data of a plurality of pixels and input weighting values in a predetermined area;
• second multiply-and-accumulate arithmetic means for digitally processing multiplication and accumulation of output image data of a plurality of pixels and output weighting values in a predetermined area;
• third multiply-and-accumulate arithmetic means for digitally processing multiplication and accumulation of state values of a plurality of pixels and state weighting values in a predetermined area;
• non-linear acting means for deciding output image data in accordance with results of processing from said first, second and third multiply-and-accumulate arithmetic means and non-linear characteristic parameters;
• parameter selecting means for selecting input weighting values used by said first multiply-and-accumulate arithmetic means, output weighting values used by said second multiply-and-accumulate arithmetic means, state weighting values used by said third multiply-and-accumulate arithmetic means, and non-linear characteristic parameters used by said non-linear acting means, wherein various types of image processing can be executed selectively in accordance with the selected input weighting values, output weighting values, state weighting values and non-linear characteristic parameters;
• and
• offset input means for entering a predetermined offset value, wherein said non-linear acting means decides output image data in accordance with results of processing from said first, second and third multiply-and-accumulate arithmetic means, said entered offset value and the non-linear characteristic parameters.

A digital image processor for processing input image data based upon a cellular neural network, comprising:

• first multiply-and-accumulate arithmetic means for digitally processing multiplication and accumulation of input image data of a plurality of pixels and input weighting values in a predetermined area;
• second multiply-and-accumulate arithmetic means for digitally processing multiplication and accumulation of output image data of a plurality of pixels and first output weighting values in a predetermined area;
• third multiply-and-accumulate arithmetic means for digitally processing multiplication and accumulation of output value data of a preceding frame image and second output weighting values;
• and
• non-linear acting means for deciding output image data in accordance with results of processing from said first, second and third multiply-and-accumulate arithmetic means and non-linear characteristic parameters.

A digital image processor for processing input image data based upon a cellular neural network, comprising:

• first multiply-and-accumulate arithmetic means for digitally processing multiplication and accumulation of input image data of a plurality of pixels and input weighting values in a predetermined area;
• second multiply-and-accumulate arithmetic means for digitally processing multiplication and accumulation of output image data of a plurality of pixels and output weighting values in a predetermined area;
• and
• non-linear acting means for deciding output image data in accordance with results of processing from said first and second multiply-and-accumulate arithmetic means and non-linear characteristic parameters;
• and
• wherein said non-linear acting means converts the input image data into the output data of 2 bits or more using non-linear characteristic parameters as defined in a step function with discontinuities having a range of -1 to 1, and converges the results of processing the range -1 to 1 of the non-linear characteristic parameters.

A digital image processor for processing input image data based upon a cellular neural network, comprising:

• first multiply-and-accumulate arithmetic means for digitally processing multiplication and accumulation of input image data of a plurality of pixels and input weighting values in a predetermined area;
• second multiply-and-accumulate arithmetic means for digitally processing multiplication and accumulation of output image data of a plurality of pixels and output weighting values in a predetermined area;
• third multiply-and-accumulate arithmetic means for digitally processing multiplication and accumulation of output state values of a plurality of pixels an state weighting values in a predetermined area;
• and
• non-linear acting means for deciding output image data in accordance with results of processing from said first, second and third multiply-and-accumulate arithmetic means and non-linear characteristic parameters.

A digital image processing method for processing input image data based upon a cellular neural network, comprising:

• a first multiply-and-accumulate arithmetic step of digitally processing multiplication and accumulation of input image data of a plurality of pixels and input weighting values in a predetermined area;
• a second multiply-and-accumulate arithmetic step of digitally processing multiplication and accumulation of output image data of a plurality of pixels and output weighting values in a predetermined area;
• and
• a non-linear acting step of deciding output image data in accordance with results of processing at said first and second multiply-and-accumulate arithmetic steps and non-linear characteristic parameters;
• and
• wherein said non-linear acting step converts the input image data into the output date of 2 bits or more using non-linear characteristic parameters as defined in a step function with discontinuities having a range of -1 to 1, and converges the results of processing to the range -1 to 1 of the non-linear characteristic parameters.