A data cache device applied to parallel filtering operation

Abstract

The invention discloses a data caching device applied to parallel filtering operation. The data caching device applied to the parallel filtering operation comprises a caching area entity, reading caching logic, initializing caching logic and updating caching logic, wherein the caching area entity is used for caching data from a multi-granularity memorizer, the reading caching logic is used for receiving reading caching area enable signals, reading caching area column numbers and reading caching area in-column excursion and reading a certain list of data of the caching area entity according to the signals, the initializing caching logic is used for initializing the caching area entity when initialization starting signals are valid, and the updating caching logic is used for bodily moving the caching area entity according to lists when list movement signals are valid and meanwhile outputting signal groups of the multi-granularity memorizer. According to the data caching device applied to the parallel filtering operation, data can be processed before filtration, data needed by the parallel filtering operation are provided in the filtration, a data locality theory is fully utilized, the number of visiting times and the number of storing times are reduced, and the maximum utilization rate of the data is achieved.

Description

technical field [0001] The invention relates to the technical field of digital signal processors and microprocessors, in particular to a data buffer structure for filtering operations, which is closely related to filtering algorithms and integrated circuit structures. Background technique [0002] In signal processing systems such as image processing, image filtering is an indispensable processing operation for obtaining images with a low signal-to-noise ratio. For different filtering templates, the calculation amount of two-dimensional filtering is different. Generally, for a K×K filtering template, calculating each filtering result requires K 2 multiplications and K 2 addition, then for an input matrix of size M×N, the amount of computation required is 2MNK 2 , so two-dimensional filtering is a computationally intensive algorithm. [0003] For the huge amount of calculation above, its data use has great local characteristics. For a K×K filter template, two adjacent filt...

AI Technical Summary

Problems solved by technology

[0008] The technical problem to be solved by the present invention is that the existing data buffering device is not suitable for parallel filtering operation, resulting in insufficient operation efficiency

Images