APPARATUS AND METHOD FOR CLASSIFICATION BASED ON CONVOLUTION NEURAL NETWORK WITH ENHANCED ACCURACY

摘要

The present invention relates to a convolutional neural network-based classification apparatus and method with improved accuracy. According to an aspect of the present invention, a convolutional neural network-based classification method with improved accuracy prepares a plurality of RGB training images according to a combination of pixels having RGB values. Step, converting the RGB values of the plurality of RGB training images into grayscale values according to a preset color conversion model to generate a plurality of grayscale images corresponding to each of the plurality of RGB training images, the plurality of grayscale images generating a histogram based on the intensity values of the elements, modifying the histogram according to a preset transformation model, and obtaining a plurality of contrast-enhanced images with improved contrast based on the modified histogram; preset learning of a plurality of contrast-enhanced images and obtaining a feature parameter by applying it to a model, and classifying a confirmation image according to a user input based on the feature parameter by selecting at least one classification value from among a plurality of preset classification values.