A multi-source optical sample generation method and system

Abstract

The present application provides a multi-source optical sample generation method and system, belonging to the technical field of image processing. In the embodiment of the present application, by receiving input parameters for generating a simulation scene of the same identification task, and according to the input parameters, a simulation scene corresponding to each optical type is constructed one-to-one, and in the simulation scene corresponding to each optical type, An optical image of this optical type and annotation information corresponding to the optical image are generated, and finally the optical image and the annotation information are used as an optical sample under this optical type. The embodiment of the present application can generate enough high-quality optical samples of various optical types that are closely related to the recognition task by means of simulation, according to the recognition requirements of the recognition task, and use the optical samples to identify the target corresponding to the recognition task. Sufficient training can improve the training effect of the target recognition model, thereby effectively improving the practical application effect of the target recognition model when performing the recognition task.

Description

technical field [0001] The present application relates to the technical field of image processing, and in particular, to a method and system for generating a multi-source optical sample. Background technique [0002] With the rapid development of deep learning technology, supervised learning methods based on deep learning technology are widely used in tasks such as target recognition, detection, and tracking in various scenarios such as environmental pollution monitoring, endangered species monitoring, and traffic monitoring. A core module when performing target recognition, detection, and tracking tasks is the target recognition model based on deep learning, and the actual application effect of the target recognition model is closely related to the training level of the target recognition model, that is, the target recognition model used in training. The quality and quantity of optical samples are related, and it is necessary to use a large number of optical samples closely...

AI Technical Summary

Problems solved by technology

[0004] This application provides a multi-source optical sample generation method and system to solve the problem of time-consuming and labor-intensive manual collection of optical samples. The number of high-quality optical samples collected is limited and the types are single, which cannot maximize the ability of the target recognition model based on deep learning technology. The problem

Images