Image encoding apparatus and method, computer program, and computer-readable storage medium

A method for controlling an image encoding computer that encodes image data for a pixel block, the method comprising:

• a first encoding step of lossily encoding input image data for the pixel block to generate encoded data;
• a second encoding step of losslessly encoding input image data for the pixel block to generate encoded data;
• and
• a selection step of, letting Lx be a code length of encoded data of the pixel block generated in the second encoding step and Ly be a code length of encoded data of the pixel block generated in the first encoding step, when Lx and Ly satisfy a predetermined non-linear boundary function f( ):
• Ly≧(Lx),
• selecting and outputting, by the image encoding computer, the encoded data generated in the second encoding step, and, when Lx and Ly do not satisfy the function f( ), selecting and outputting, by the image encoding computer, the encoded data generated in the first encoding step,
• wherein, when an X-Y coordinate space is defined in which the X axis represents the code length of a pixel block encoded in said second encoding step and the Y axis represents the code length of the pixel block encoded in said first encoding step, and
• in the X-Y coordinate space, an area in which encoded blocks of a predetermined text image are plotted is defined as the distribution area “T”, an area in which encoded blocks of a predetermined computer-graphic image whose gradation is changed smoothly are plotted is defined as the distribution area “G”, and an area in which encoded blocks of a predetermined natural image are plotted is defined as the distribution area “I”,
• the predetermined non-linear boundary function f( ) has, in the X-Y coordinate space, a curve portion which extends from the origin point and passes between the distribution areas “G”and “I”, and a linear portion which passes between the distribution areas “T” and “I”.

A computer-readable storage medium having stored thereon a computer program that, when read and executed by a computer, causes the computer to function as an image encoding apparatus for encoding image data for a pixel block, wherein said computer program causes the computer to execute:

• a first encoding step of lossily encoding input image data for the pixel block to generate encoded data;
• a second encoding step of losslessly encoding input image data for the pixel block to generate encoded data;
• and
• a selection step of, letting Lx be a code length of encoded data of the pixel block generated in said second encoding step and Ly be a code length of encoded data of the pixel block generated in said first encoding step, when Lx and Ly satisfy a predetermined non-linear boundary function f( )
• Ly≧f(Lx),
• selecting and outputting the encoded data generated in said second encoding step, and, when Lx and Ly do not satisfy the function f( ), selecting and outputting the encoded data generated in said first encoding step,
• wherein, when an X-Y coordinate space is defined in which the X axis represents the code length of a pixel block encoded in said second encoding step and the Y axis represents the code length of the pixel block encoded in said first encoding step, and
• in the X-Y coordinate space, an area in which encoded blocks of a predetermined text image are plotted is defined as the distribution area “T”, an area in which encoded blocks of a predetermined computer-graphic image whose gradation is changed smoothly are plotted is defined as the distribution area “G”, and an area in which encoded blocks of a predetermined natural image are plotted is defined as the distribution area “I”,
• the predetermined non-linear boundary function f( ) has, in the X-Y coordinate space, a curve portion which extends from the origin point and passes between the distribution areas “G” and “I”, and a linear portion which passes between the distribution areas “T” and “I”.