Data compression apparatus and data compression method
AssigneeFujitsu Limited
InventorYoshiaki Ogasawara
Life of this patent
Estimating the term…
Abstract
A processor calculates, from image data, a plurality of first coefficients corresponding to different frequencies using a Fourier transform and stores first frequency domain data, from which first coefficients whose magnitude is below a threshold have been excluded, in a memory. The processor also calculates, from kernel data, second coefficients of frequencies corresponding to the first coefficients indicated by the first frequency domain data using a Fourier transform and stores second frequency domain data in the memory. The processor uses the first frequency domain data, the second frequency domain data, and an inverse Fourier transform to generate a feature map indicating a result of applying the kernel data to the image data.
Claims7
A data compression apparatus comprising:
- a memory that stores image data and kernel data to be applied to the image data;
- and
- a processor that executes a process including:
- calculating, from the image data, a plurality of first coefficients corresponding to different frequencies using a Fourier transform;
- storing first frequency domain data, which indicates a subset of the plurality of first coefficients from which one or more first coefficients of a magnitude below a threshold have been excluded, in the memory;
- calculating, from the kernel data, second coefficients of frequencies corresponding to the subset of the plurality of first coefficients indicated by the first frequency domain data using a Fourier transform;
- storing second frequency domain data indicating the second coefficients in the memory;
- and
- generating a feature map indicating a result of applying the kernel data to the image data using the first frequency domain data, the second frequency domain data, and an inverse Fourier transform.
A data compression method comprising:
- calculating, by a processor, a plurality of first coefficients corresponding to different frequencies from image data using a Fourier transform, and storing first frequency domain data, which indicates a subset of the plurality of first coefficients from which one or more first coefficients of a magnitude below a threshold have been excluded, in a memory;
- calculating, by the processor, second coefficients of frequencies corresponding to the subset of the plurality of first coefficients indicated by the first frequency domain data from kernel data, which is to be applied to the image data, using a Fourier transform, and storing second frequency domain data indicating the second coefficients in the memory;
- and
- generating, by the processor, a feature map, which indicates a result of applying the kernel data to the image data, using the first frequency domain data, the second frequency domain data, and an inverse Fourier transform.
A non-transitory computer-readable storage medium storing a computer program that causes a computer to execute a process comprising:
- calculating a plurality of first coefficients corresponding to different frequencies from image data using a Fourier transform, and storing first frequency domain data, which indicates a subset of the plurality of first coefficients from which one or more first coefficients of a magnitude below a threshold have been excluded, in a memory of the computer;
- calculating second coefficients of frequencies corresponding to the subset of the plurality of first coefficients indicated by the first frequency domain data from kernel data, which is to be applied to the image data, using a Fourier transform, and storing second frequency domain data indicating the second coefficients in the memory;
- and
- generating a feature map, which indicates a result of applying the kernel data to the image data, using the first frequency domain data, the second frequency domain data, and an inverse Fourier transform.
Dependencies are inferred from the claim text.