The present invention comprises the steps of: constructing a multi-dimensional neural network using an auxiliary neural network that learns a main network and learns about images that are not recognized by the main network; Regenerating data to be used as an input of a decision-making neural network using the output result of the multidimensional neural network; And constructing and learning a decision-making neural network to input the regenerated data; and an image recognition method through deep learning learning of a multi-dimensional neural network and a decision-making neural network, and a system using the same. According to the present invention, by using a multidimensional neural network and a decision-making neural network, it has the effect of shortening the learning and execution time, which has emerged as a problem of the existing CNN (Convolution Neural Network), and improving the accuracy.
展开▼