matlab-基于CNN卷积神经网络的目标分类训练和测试matlab仿真-源码

clc; clear; close all; warning off; addpath(genpath(pwd)); rng('default'); digitDatasetPath = fullfile('NewData'); digitData = imageDatastore(digitDatasetPath,'IncludeSubfolders',true,'LabelSource','foldernames'); figure; perm = randperm(3000,20); for i = 1:20 subplot(4,5,i); imshow(digitData.Files{perm(i)}); end trainNumFiles = 0.85; [trainDigitData,valDigitData] = splitEachLabel(digitData,trainNumFiles); %Define the convolutional neural network architecture. layers = [ imageInputLayer([48 12 1]) convolution2dLayer(3,16,'Padding',1) batchNormalizationLayer reluLayer maxPooling2dLayer(2,'Stride',2) %dropoutLayer(0.5) convolution2dLayer(3,32,'Padding',1) batchNormalizationLayer reluLayer maxPooling2dLayer(2,'Stride',2) %dropoutLayer(0.5) convolution2dLayer(3,64,'Padding',1) batchNormalizationLayer reluLayer maxPooling2dLayer(2,'Stride',2) convolution2dLayer(3,128,'Padding',1) batchNormalizationLayer reluLayer fullyConnectedLayer(9) softmaxLayer classificationLayer]; options = trainingOptions('sgdm',... 'MaxEpochs',20, ... 'ValidationData',valDigitData,... 'ValidationFrequency',30,... 'Verbose',false,... 'Plots','training-progress'); net = trainNetwork(trainDigitData,layers,options); predictedLabels = classify(net,trainDigitData); valLabels = trainDigitData.Labels; accuracy = sum(predictedLabels == valLabels)/numel(valLabels) *100 %test digitDatasetPath = fullfile('NewData'); digitData = imageDatastore(digitDatasetPath,'IncludeSubfolders',true,'LabelSource','foldernames'); real_labels=digitData.Labels; figure; perm = randperm(3000,20); for i = 1:20 subplot(4,5,i); imshow(digitData.Files{perm(i)}); realL=real_labels(perm(i)); I=imread(digitData.Files{perm(i)}); predictedLabels = classify(net,I); Lidx=double(predictedLabels); Label={'1','2','3','4','5','6','7','8','9'}; title(['CNN=',Label{Lidx}]) ylabel(['Real=',Label{realL}]) end