Candidatus Liberibacter spp. online rapid detection system and method based on deep learning

Abstract

A Candidatus Liberibacter spp. online rapid detection system based on deep learning includes a mobile phone client and a server end; the server end includes a data transmission module, a leaf and fruit detection module, a Candidatus Liberibacter spp. diagnosis module, and a sample database; a model of the leaf and fruit detection module is trained by using the data with instance annotations in a sample database to obtain the optimal model; a model of the Candidatus Liberibacter spp. diagnosis module is trained by using the data with diagnostic annotations in the sample database to obtain the optimal model; the leaf and fruit detection module transmits the output and detected single leaf and single fruit images to the Candidatus Liberibacter spp. diagnosis module. A Candidatus Liberibacterspp. online rapid detection method based on deep learning adopts the Candidatus Liberibacter spp. online rapid detection system based on deep learning. The invention belongs to the technical field ofintelligent identification of Candidatus Liberibacter spp., and has the advantages of rapid, efficient and reliable detection.

Description

technical field [0001] The invention belongs to the technical field of intelligent identification of citrus huanglongbing, and in particular relates to an online rapid detection system and method for citrus huanglongbing based on deep learning. Background technique [0002] Citrus is one of the fruits with the largest production volume in the world, and it is also one of the largest fruit varieties planted in southern my country. It plays a very important role in the agricultural economy. And citrus Huanglongbing (HLB) is destructive to the production of citrus. The disease causes citrus trees to show symptoms such as mottled leaves, yellow leaves, weak tree vigor, red-nosed fruit or green fruit that does not turn color, etc., and has the potential to spread. The characteristics of fast speed and great harm. Once the citrus trees are infected with the disease, the light ones will seriously affect the fruit yield and quality, and the severe ones will cause the citrus plants ...

AI Technical Summary

Problems solved by technology

At present, the most reliable detection method for HLB is PCR detection technology, but the detection process of this method is cumbersome, the cycle is long, the detection cost is high, and the detection environment and operation requirements are high, which limits the application of this method in actual production.

In addition, methods such as graft diagnosis, serological diagnosis, field diagnosis, and DNA probe hybridization are difficult to promote in actual production due to low diagnostic accuracy, long time-consuming, high cost, and cumbersome processes.

Images