A scalable in-memory image processing system and method based on WHT

By splitting the image into data groups and mapping the WHT operator matrix using a memristor array, the problem that fixed-size memristor arrays are difficult to adapt to WHT operations of different scales is solved, and scalable in-memory image processing with low power consumption and low latency is achieved.

CN119784600BActive Publication Date: 2026-06-02HUAZHONG UNIV OF SCI & TECH

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HUAZHONG UNIV OF SCI & TECH
Filing Date
2024-12-20
Publication Date
2026-06-02

Smart Images

  • Figure CN119784600B_ABST
    Figure CN119784600B_ABST
Patent Text Reader

Abstract

The application belongs to the technical field of signal analysis and signal processing, and specifically discloses an extensible in-memory image processing system and method based on WHT. According to the application, a first input processing module is used to split an image into a plurality of first data groups in a first direction, and output corresponding first voltage pulse signals. An operation unit is used to process the first voltage pulse signals corresponding to each first data group by using a memristor array to map a matrix corresponding to a WHT operator, and obtain a plurality of first current pulse signals corresponding thereto. An output processing module is used to convert the first current pulse signals output by the operation unit into corresponding second data groups, recombine the plurality of second data groups in a second direction, and obtain final output data. According to the application, a fixed-size memristor array can be used to realize WHT operation of any scale, thereby reducing the operation amount of the system while maintaining low power consumption.
Need to check novelty before this filing date? Find Prior Art