Method of compressing digital images

A method for compressing a digital image comprising a matrix of elements, each element comprising at least one digital component for representing a pixel, the method comprising:

• splitting the digital image into a plurality of blocks, and calculating for each block a group of discrete cosine transform (DCT) coefficients for the different types of components;
• quantizing the DCT coefficients for each group using a corresponding quantization table scaled by a gain factor for achieving a target compression factor;
• further quantizing the DCT coefficients for each group using the corresponding quantization table scaled by a pre-set factor;
• arranging the further quantized DCT coefficients in a vector;
• calculating a basic compression factor provided by the quantization table scaled by the pre-set factor as a first function of the vector by
• determining a first number of bits required to encode the vector for each block, and
• summing all the first numbers of bits associated with each block with a second number of bits required to encode control values, and dividing the sum by a number of elements of the digital image;
• and
• estimating the gain factor as a second function of the basic compression factor.

A device for compressing a digital image comprising a matrix of elements, each element comprising at least one digital component for representing a pixel, the device comprising:

• discrete cosine transform (DCT) means for splitting the digital image into a plurality of blocks, and calculating for each block a group of DCT coefficients for the different types of components;
• quantization means for
• quantizing the DCT coefficients of each group using a corresponding quantization table scaled by a gain factor for achieving a target compression factor, and
• further quantizing the DCT coefficients of each group using the corresponding quantization table scaled by a pre-set factor;
• arranging means for arranging the further quantized DCT coefficients in a vector;
• calculation means for calculating a basic compression factor provided by the quantization table scaled by the pre-set factor as a first function of the vector, the calculating comprising
• determining a first number of bits required to encode the vector for each block,
• summing all the first numbers of bits associated with each block with a second number of bits required to encode control values, and
• dividing the sum by a number of elements of the digital image;
• and
• estimation means for estimating the gain factor as a second function of the basic compression factor, the second function being determined experimentally according to the target compression factor.

A digital still camera comprising:

• an image acquisition unit for transmitting light corresponding to an image of scene;
• a sensor unit connected to said image acquisition unit for providing digital image of scene, the digital image comprising a matrix of elements, each element comprising at least one digital component for representing a pixel;
• and
• a control device for compressing the digital image and comprising
• a discrete cosine transform (DCT) unit splitting the digital image into a plurality of blocks, and calculating for each block a group of DCT coefficients for the different types of components;
• a quantization unit for
• quantizing the DCT coefficients of each group using a corresponding quantization table scaled by a gain factor for achieving a target compression factor, and
• further quantizing the DCT coefficients of each group using the corresponding quantization table scaled by a pre-set factor;
• a zig-zag unit for arranging the further quantized DCT coefficients a vector;
• and
• a processor for calculating a basic compression factor provided by the quantization table scaled by the pre-set factor as a first function of the vector, and for estimating the gain factor as a second function of the basic compression factor, the second function being determined experimentally according to the target compression factor, and the calculating comprising
• determining a first number of bits required to encode the vector for each block,
• summing all the first numbers of bits associated with each block with a second number of bits required to encode control values, and
• dividing the sum by a number of elements of the digital image.