Standardization platform for machine vision algorithm framework

Abstract

The invention relates to a standardization platform for a machine vision algorithm framework which comprises data processing units, wherein the data processing units are configured at different PC terminals, and data transmission is carried out between the data processing units and between the data processing units and clients through communication modules; each data processing unit comprises the algorithm framework, and the algorithm framework is used for carrying out algorithm flow processing on data output by the data acquisition equipment and outputting a detection result; and when the associated output data of a certain configured algorithm is used as associated input by the algorithms under other data processing units, the algorithm framework is realized by adopting a cross-PC calling function. The standardization process of algorithm development and parameter configuration is enhanced, the algorithm development difficulty is reduced, a large amount of repeated development work is avoided, and the project development efficiency is improved; algorithm data association under a complex visual structure is supported, and flexibility and convenience are provided for implementation of complex projects.

Description

technical field [0001] The invention relates to the technical field of visual inspection, in particular to a machine vision algorithm frame standardization platform. Background technique [0002] The main functions of the algorithm integration framework are: (1) to provide a standardized algorithm integration platform for the rapid integration and verification of various functional algorithms; (2) as a functional platform that can connect clients, hardware devices and algorithms to realize various data analysis and delivery. [0003] Existing algorithm integration frameworks generally include four major processes: algorithm library import, template creation, algorithm execution, and result display. Algorithm import refers to the design of parameter configuration and dynamic library development that meet the requirements of the corresponding algorithm framework. Algorithm task processes are assigned to each data by creating templates. The framework calls and executes each al...

AI Technical Summary

Problems solved by technology

[0004] However, the existing algorithm integration framework does not standardize the configuration and integration of algorithm parameters to a high degree. The parameter configuration process is complicated and time-consuming. Algorithm developers need to analyze their own algorithm parameters.

[0005] Moreover, the data information between different algorithms is independent of each other, so complex data structures and complex association templates cannot be shared, because the framework does not have a transmission and synchronization mechanism for data under different PCs

Images