Plant nematode detection method and system based on multi-feature combination based on deep learning

Abstract

The invention discloses a plant nematode detection method and system based on deep learning and multi-feature combination. The present invention uses the first deep learning neural network to extract the key features of plant nematodes, and respectively detects the key characteristic regions such as the head, female tail, and male tail of the plant nematodes; and uses the second deep learning neural network as the identification and detection network to achieve Single-feature detection, dual-feature joint detection and three-feature joint detection have high detection accuracy. Finally, a humanized nematode detection interface is designed, and an intelligent nematode detection system is completed. The nematode detection and identification scheme can be selected according to the degree of difficulty in distinguishing nematodes. The invention adds a mechanism for effectively extracting key features, and a method for multi-feature joint classification. The invention can accurately detect plant nematode data with small key feature points and are difficult to distinguish, and utilizes multi-feature combination to realize hierarchical detection of pine wood nematodes and pine wood nematodes.

Description

technical field [0001] The invention relates to the technical field of deep learning image intelligent detection and retrieval, in particular to a method and system for intelligent detection of plant nematodes based on a combination of deep learning and multi-features. Background technique [0002] Affected by alien species, the spread of pine wood nematode is becoming more and more serious in my country, resulting in huge economic losses every year. It is imperative to increase quarantine efforts, improve the ability to intercept the epidemic from the source, and curb the spread of pine wood nematode. However, the accurate detection of B. xylophilus among various nematodes with very similar characteristics presents a great challenge to the technicians of the detection agency. How to efficiently and accurately detect B. xylophilus has always been a very critical problem. Due to the small size of nematodes, the key features of nematodes such as head and tail are smaller and ...

AI Technical Summary

Problems solved by technology

Due to the small size of the nematode, the key feature points of the nematode such as the head and tail are smaller and less obvious than the nematode body. Sometimes it is necessary to jointly detect multiple features to distinguish the type of B. xylophilus, which is very difficult to identify.

[0003] The traditional detection method is based on the shape of nematodes, mainly relying on accumulated experience to identify familiar groups, which is greatly limited by manpower

Images