Multi-algorithm framework compatible processing method and system based on machine vision

The invention discloses a multi-algorithm framework compatible processing method and system based on machine vision. The method comprises the following steps: acquiring a target algorithm framework; wherein the target algorithm framework comprises one or more of a caffe algorithm framework, an ONNX algorithm framework, a DarkNet algorithm framework and a Pytorch algorithm framework; analyzing and quantifying the target algorithm framework to generate an instruction and a file required by an FPGA (Field Programmable Gate Array); instructions and files required by the FPGA are transmitted to the FPGA; and the FPGA calls a corresponding operator for operation according to the received instruction and file to obtain an operation result. According to the method, a multi-algorithm framework and multi-network model can be supported, and the problem of inflexible application caused by the fact that part of methods and systems only support single algorithm framework operation in the prior art is solved. 本发明公开了一种基于机器视觉