Establishment method and application of strawberry identification and counting model

Abstract

The invention provides a construction method and application of a strawberry identification and counting model, and the method comprises the following steps: firstly, obtaining a strawberry image set which is derived from a plurality of strawberry images shot in a field environment; preprocessing the acquired strawberry image set and constructing a data set; the preprocessing process comprises image size adjustment, image amplification and image annotation; based on a Faster RCNN model, improvement is carried out, and a strawberry identification and counting model is built; the strawberry identification and counting model comprises four main modules, namely a feature extraction module, a region suggestion module, a frame regression and classification module and a strawberry counting module; and finally, dividing the data set, defining the data set as a training set, a verification set and a test set, training the established strawberry identification and counting model, and completing the test. The strawberry identification and counting model provided by the invention has good detection performance and counting capability, and can provide theoretical basis and technical support for automatic strawberry picking and yield estimation.

Description

technical field [0001] The invention belongs to the technical field of strawberry identification and counting, and in particular relates to a construction method and application of a strawberry identification and counting model. Background technique [0002] Strawberries are one of the most widely grown fruits in the world. In China, strawberry has become an important economic fruit crop, and its annual output ranks first in the world. The European strawberry market is also in great demand. To meet this demand, many EU and non-EU companies are expanding their acreage. With the rapid development of the global strawberry planting industry, strawberry picking is faced with the problem of labor shortage due to its heavy dependence on labor. In Japan, strawberry picking accounts for a quarter of all working hours in the strawberry industry; in Norway, strawberry picking costs account for more than 40-60% of total labor costs. In China, the aging of the strawberry labor force ...

AI Technical Summary

Problems solved by technology

[0004] The above studies have solved the problems of low strawberry recognition rate and difficulty in strawberry fruit recognition in complex environments, but there is no combination of target recognition, complex environments, mature and immature recognition and strawberry fruit counting. Research

Although the recognition rate of mature strawberries is already relatively high, there are still great challenges in the recognition accuracy of immature strawberries, which need to be further improved.

Images