Application Of Image Recognition Based On The Determination Of Element Imbalance In Seedling Stage Of Millet

With the increasing planting area of millet in China,millet has become an important crop of crops in northern China and even in the whole country.During the seedling stage and growth process of millet,it is necessary to detect the soil and fertilization environment at all times to avoid large-scale deficiency of potassium and magnesium.Accurately identifying whether there is element loss in the seedling stage and growth process is important to timely supplement the required nutrients for the millet and predict the future growth of the millet.With the development of agricultural information technology and the good effect of computer technology in image recognition,these provide strong support for the identification of whether the images in the growth of millet lack of element.The deep learning method has the effect that the traditional image recognition method is difficult to achieve in the field of image recognition.With the emergence and continuous improvement of the artificial neural network model,the recognition accuracy and accuracy of the deep learning method in the field of image recognition are continuously improved,which provides an important basis for the research topics in this paper.This paper proposes an improved convolutional neural network.The Inception structure is used to optimize the model,meanwhile,combined depth features at different scales to identify variant plaque patterns of various sizes.Then the network model can accurately extract the features of the input valley graph,and can further improve the recognition and classification accuracy of the network.The experimental results show that the proposed method can achieve higher recognition accuracy in effectively identifying the missing element types of the millet images.All these results can further verify the validity of the research topic and the efficiency of the proposed method.