An automatic identification method of apple phenology by fusing image and time features

Abstract

The invention belongs to the technical field of computer vision, and in particular relates to an automatic recognition method of apple phenology by fusing images and time features, comprising the following steps: using a fully connected network to abstract a one-dimensional time feature vector, using a convolutional neural network to extract The object features are extracted, the one-dimensional time features and the abstract image features are fused in series, and the fully connected layer is used to classify the fused feature vectors to identify the phenological period to which the image belongs. The present invention first proposes a scientific acquisition and processing method for apple tree images, uses image random cropping and information deletion methods to expand data for the problem of small data volume, and fuses time features on the basis of convolution network extraction of apple tree image features to obtain A more representative feature map is used for the calculation of the probability of the phenological category, which improves the accuracy of the phenological identification of apple trees. The invention is used for identifying the phenophase of apples.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to an automatic identification method of apple phenology by fusing images and time features. Background technique [0002] Agriculture is a basic industry in our country. The output of apples in our country is huge, and it is the main economic crop in many places. With the help of scientific and technological innovation, it is helpful to promote the scientific and standardization of apple planting, which is crucial to improving the productivity of apple orchards, simultaneously reducing the cost of apple tree planting and increasing the economic benefits it brings. [0003] The phenological period refers to the natural phenomenon that occurs on a yearly basis when plants are affected by biotic and abiotic factors. The growth cycle of an apple tree is long, and it needs to experience bud expansion, bud opening, leaf expansion, flowering, tipping, fruiting and leaf...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the above-mentioned technical problems of large uncertainty, time-consuming and labor-intensive, and high cost of the existing methods for judging the phenology of fruit trees in apple orchards, the present invention provides a fusion image and time feature with high efficiency, high accuracy, and low cost. Automatic identification method of apple phenology

Images