Plant nematode data automatic labeling and classification recognition method based on deep learning

Abstract

The invention discloses a plant nematode data automatic annotation and classification recognition method based on deep learning, and the method comprises the steps: training a deep learning network through a small number of manually annotated plant nematode images, adding attention loss to the deep learning network, thereby strengthening a plant nematode annotation region, then using the network weight of the model to label a large amount of non-artificially labeled plant nematode image data, completing automatic labeling of a plant nematode identification project data set through a pluralityof plant nematode image data labeling and training processes, and obtaining an enhanced model corresponding to the plant nematode identification project at the same time. According to the method, theplant nematode data can be effectively labeled, so that the training efficiency of the plant nematode data and the accuracy of plant nematode classification and recognition are improved.

Description

technical field [0001] The invention relates to the technical field of deep learning image intelligent labeling and retrieval, in particular to a deep learning network-based intelligent labeling and classification recognition method for plant nematode images. Background technique [0002] In recent years, artificial intelligence and big data have become the focus of attention in various fields at home and abroad. Among them, the deep learning algorithm that has been widely concerned has been successfully applied to various fields such as speech recognition and graphic recognition. With the continuous deepening of various industries Informatization construction has accumulated a large amount of data in various fields such as medical treatment, biology, finance, and law. How to efficiently manage and organize these image data, and how to effectively label these data has become a pain point in the industry. For some less common deep learning data sets, such as the plant nematod...

AI Technical Summary

Problems solved by technology

The existing deep learning labeling methods are mainly manual labeling, but manual labeling is not only time-consuming and labor-intensive, but also consumes a lot of manpowe

Images