MATLAB水果识别代码

以下是MATLAB实现水果识别的示例代码：

1. 数据准备

% 加载数据集
fruitDataset = imageDatastore('fruits', 'IncludeSubfolders', true, 'LabelSource', 'foldernames');

% 将数据集分成训练集和测试集
[trainDataset, testDataset] = splitEachLabel(fruitDataset, 0.8, 'randomize');

% 显示部分数据
numTrainImages = numel(trainDataset.Labels);
idx = randperm(numTrainImages, 4);
figure
for i = 1:4
    subplot(2,2,i)
    I = readimage(trainDataset, idx(i));
    imshow(I)
end

2. 特征提取

% 提取训练集的图像特征
bag = bagOfFeatures(trainDataset);

% 显示部分特征
figure
bar(bag.VocabularyCounts)
title('Visual word occurrences')
xlabel('Visual word index')
ylabel('Frequency of occurrence')

3. 训练分类器

% 提取训练集的特征向量和标签
trainFeatures = encode(bag, trainDataset);
trainLabels = trainDataset.Labels;

% 训练SVM分类器
classifier = fitcecoc(trainFeatures, trainLabels, 'Learners', 'Linear', 'Coding', 'onevsall', 'ObservationsIn', 'columns');

4. 测试和评估

% 提取测试集的特征向量和标签
testFeatures = encode(bag, testDataset);
testLabels = testDataset.Labels;

% 使用分类器对测试集进行分类
predictedLabels = predict(classifier, testFeatures);

% 计算分类器的准确率、召回率、F1值等指标
accuracy = mean(predictedLabels == testLabels);
confMat = confusionmat(testLabels, predictedLabels);
precision = diag(confMat)./sum(confMat,2);
recall = diag(confMat)./sum(confMat,1)';
f1 = 2*(precision.*recall)./(precision+recall);
meanF1 = mean(f1);

这是一个简单的水果识别示例，实际应用中可能需要更复杂的特征提取方法和分类器。