Quantization step size adjusting circuit using edge detections

A circuit for adjusting quantization step sizes in sampling of an image signal to provide a digital signal, said circuit comprising:

• means for receiving said image signal on a frame-by-frame basis with each frame of said image signal comprised of a plurality of image blocks;
• means for determining a quantized step adjusting signal representative of a quantized step size for quantization of said image signal in dependence upon a determination of edge values representative of peripheral pixels of respective ones of a plurality of addresses in each frame of said image signal designated by control signals and a determination of a selected scale factor signal in accordance with pixel density representation of each image block comprised of a predetermined size of pixels in each frame, said quantized step adjusting signal determining means comprising:
• edge detector means for detecting said edge values representative of peripheral pixels of respective ones of said plurality of addresses in each frame of said image signal designated by said control signals;
• frame memory means for storing detected edge values for each frame of said image signal;
• block format means for accumulating stored edge values from said frame memory means for each image block in each frame of said image signal in accordance with said control signals to provide an accumulated edge value signal;
• scaling factor means for determining scaling factor values for said accumulated edge value signal in accordance with said pixel density representation of each image block to provide said selected scaling factor signal;
• and
• means for providing said quantized step adjusting signal in dependence upon said selected scaling factor signal;
• means for quantizing each image block in each frame of said image signal in accordance with said quantized step adjusting signal to provide a quantized signal;
• and
• means for encoding said quantized signal to provide said digital signal.

A circuit for adjusting quantization step sizes of an image signal, comprising:

• means coupled to receive said image signal representative of a picture frame, for dividing said image signal into a plurality of image blocks each comprised of a predetermined number of pixels;
• edge detector means for detecting edge values representative of peripheral pixels of respective ones of a plurality of addresses of said image signal designated by control signals;
• frame memory means for storing detected edge values for each picture frame;
• block format means for accumulating stored edge values from said picture frame for each image block of said plurality of image blocks in accordance with said control signals to provide an accumulated edge value signal;
• scaling factor means for determining scaling factor values for said accumulated edge value signal in accordance with pixel density representation of each image block of said image signal to provide a selected scaling factor signal;
• means for providing said quantized step adjusting signal in dependence upon said selected scaling factor signal;
• means for quantizing each image block of said image signal in accordance with said quantized step adjusting signal to provide a quantized signal;
• and
• means for encoding said quantized signal to provide a digital signal.

A device for adjusting quantization step sizes, comprising:

• means for receiving an image signal;
• edge detector means for extracting contour data for each frame of said image signal upon detection of edge values of corresponding addresses in each frame of said image signal;
• frame memory means for storing the contour data on a frame-by-frame basis;
• block format means for dividing the contour data for each frame of said image signal from said frame memory means into a plurality of data blocks having a predetermined size of pixels, and for generating an accumulated edge value signal by accumulating edge values of each data block of said image signal;
• and
• scaling factor generator means for generating a selected one of a plurality of scaling factors in accordance with pixel density representation of each data block in response to the accumulated edge value signal, said scaling factor generator means comprising:
• comparator means for generating said plurality of scaling factors in dependence upon a comparison of said accumulated edge value signal with a plurality of reference values;
• and
• means for generating said selected one of said plurality of scaling factors by encoding said plurality of scaling factors in dependence upon said pixel density representation of each data block of said image signal.

An image compression method, comprising the steps of:

• receiving an image signal on a frame-by-frame basis;
• providing a plurality of pixel values representative of peripheral pixels of corresponding ones of a plurality of addresses in each frame;
• providing a plurality of multiple pixel values by multiplying said plurality of pixel values with a constant value;
• separating a negative value portion and a positive value portion from said plurality of multiple pixel values;
• and
• extracting contour data for each frame of said image signal upon detection of edge values of corresponding ones of said plurality of addresses in each frame of said image signal by combining said negative value portion and said positive value portion;
• temporarily storing the contour data on said frame-by-frame basis onto a memory;
• dividing the contour data for each frame of said image signal from said memory into a plurality of data blocks having a predetermined size of pixels;
• generating an accumulated edge value signal by accumulating edge values of each data block of said image signal;
• generating a plurality of scaling factors in dependence upon a comparison of said accumulated edge value signal with a plurality of reference values;
• and
• generating a selected one of said plurality of scaling factors by encoding said plurality of scaling factors in dependence upon pixel density of each data block of said image signal.

An image compression device, comprising:

• means for receiving an image signal representative of an image;
• means for detecting edge values representative of peripheral pixels of corresponding ones of a plurality of addresses of said image signal and for formatting said image signal from each data frame into a plurality of data blocks each comprised of a predetermined number of pixels, said edge values detecting and data formatting means comprising:
• means for generating a plurality of latched pixel values by latching said image signal defined by a plurality of addresses respectively representative of peripheral pixels in accordance with control signals;
• means for generating a plurality of multiple pixel values by multiplying said plurality of latched pixel values with a constant value;
• means for generating a negative discrete value and a positive discrete value by adding and separating said plurality of multiple pixel values;
• means detecting said edge values representative of said peripheral pixels corresponding to said plurality of addresses of said image by subtracting said negative discrete value from said positive discrete value;
• memory means for storing the detected edge values for each data frame;
• and
• means for accumulating said detected edge values of each data block of said data frame and dividing said image signal from each data frame into said plurality of data block;
• means for generating a plurality of scaling factors respectively representative of the state of density distribution of each data block of said data frame;
• means for generating a quantizing step size value corresponding to each scaling factor for each data block of said data frame;
• and
• means for quantizing said image signal in accordance with said quantizing step size value.

A quantization step size circuit, comprising:

• means for receiving image data on a frame-by-frame basis;
• edge detector means for successively detecting edge values representative of peripheral pixels of respective addresses of said image data in each frame;
• frame memory means for storing detected edge values for each frame of said image data comprised of a plurality of sub-blocks;
• means for accumulating stored edge values for each sub-block of said each frame successively to provide an accumulated edge value for each sub-block of said image data;
• scaling factor means for responding to said accumulated edge value for each sub-block of said image data to generate a selected scaling factor based upon pixel density representation of each sub-block of said image data, said scaling factor means comprising:
• first comparing means for making a first comparison between said accumulated edge value for each sub-block of said image data in each frame with a first reference value, and for generating one of a first scaling factor, a first compared signal, and a second compared signal in dependence upon said first comparison;
• first logic means for responding to said first and second compared signals and generating a first logic signal;
• second comparing means for making a second comparison between said accumulated edge value for each sub-block of said image data in each frame with a second reference value, and for generating one of third, fourth, and fifth compared signals in dependence upon said second comparison;
• second logic means for logically responding to said third and fourth compared signals and generating a second logic signal;
• third logic means for locally responding to said first logic signal and said fifth compared signal and generating a second scaling factor;
• third comparing means for making a third comparison between said accumulated edge value for each sub-block of said image data in each frame with a third reference value, and for generating one of sixth, seventh, and eighth compared signals in dependence upon said third comparison;
• fourth logic means for logically responding to said second logic signal and said eighth compared signal and generating a third scaling factor;
• fifth logic means for logically responding to said sixth and seventh compared signals and generating a fourth scaling factor;
• and
• encoder means for encoding said first, second, third, and fourth scaling factors to produce said selected scaling factor signal for quantization of each sub-block of said image data;
• and
• means for quantizing each sub-block of said image data in said each frame based upon said selected scaling factor, and for encoding quantized image data to generate digital image data.