Big Data Analysis for Multispectral Images Recognition Based on Deep Learning
Serhii Voloshyn, Роман Пелещак, Іван Пелещак, Victoria Vysotska · 2021 IEEE 16th International Conference on Computer Sciences and Information Technologies (CSIT) · 2021
The scientific article considers multiclass classification, which is used to solve problems of recognising multispectral images. In particular, we are solving this problem using a neural network model, which is created using the Python programming language with the TensorFlow library. The article also describes the stages of establishing and optimal parameters of the model. The topical problem of multispectral image recognition is investigated. Neural networks are used for the solution. The neural network model, its number of layers, activator functions, optimiser and number of neurons in each layer were successfully selected. An accuracy of 97% is achieved for this application Using these optimal parameters.