An image processing method based on dual-mode memristive bridge synaptic circuit

Abstract

The invention provides an image processing method based on a dual-mode memristor bridge synapse circuit, comprising: constructing a double-mode memristor bridge synapse circuit through linear and nonlinear memristor bridge synapse circuits and an output memristor bridge synapse circuit based on positive and negative pulses. Mode memristive bridge synapse circuit; simulate weights according to the dual-mode memristive bridge synapse circuit, and construct a dual-mode memristive bridge neural network by simulating weights; combine the dual-mode memristive bridge synapse neural network with the cellular neural network Combined, and extract the edge of the input image through the combined neural network to identify the target element in the input image. The beneficial effects of the present invention are that the response time and weight of the dual-mode memristive bridge synaptic neural network of the present invention are faster and more accurate than the traditional ones. Therefore, dual-mode memristive bridge synaptic neural networks are expected to be more real-time and address more complex image processing.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image processing method based on a dual-mode memristive bridge synapse circuit. Background technique [0002] Artificial neural network has always been a research hotspot. Since there are hundreds of millions of neurons and synapses in the human brain, implementing synaptic circuits is important for building brain-like machines. In artificial neural networks, a large number of research results show that memristors can be used to simulate artificial synapses. Furthermore, it has great potential due to the properties of memory synaptic circuits. Among them, the most widely studied analog synaptic structure is the memory bridge structure, which has the advantages of simple structure, precise control and high integration. [0003] But in the prior art: [0004] "Kim Hyun-sook et al. (2011). Pulse-based memristor circuits for neural synapse weighting. IEEE Transactions...

AI Technical Summary

Problems solved by technology

However, it uses a synaptic bridge circuit consisting of five identical memristors, which cannot be represented by a value of 0 for synaptic weights

[0006] "Chen Jing (2018). Neural network and its application based on memristor electronic synapses. [Master's thesis]. [China]: Southwest University." proposed a method consisting of three identical memristors and two MOS The bridge synapse circuit composed of transistors makes up for the deficiency that the weight of the synapse circuit composed of five identical memristors cannot be zero, but it cannot suppress the nonlinearity of memristors

[0007] For the above defects, there is no solution for the time being

Images