高光谱图像分类

本次作业的主要内容为：高光谱分类，注意力机制
阅读论文《HybridSN: Exploring 3-D–2-DCNN Feature Hierarchy for Hyperspectral Image Classification》，思考3D卷积和2D卷积的区别。阅读代码：https://gaopursuit.oss-cn-beijing.aliyuncs.com/202010/AIexp/10%20-%20HybridSN_GRSL2020.mhtml

思考问题一：3D卷积和2D卷积的区别

3D卷积比2D卷积多一个深度信息，2D卷积可以认为是深度为1的3D卷积，在PyTorch网络骨架里的区别是，2D卷积 shape为(batch_size,channel,height,weight)，3D卷积shape为(batch_size,channel,depth,height,weight)。
二维卷积：一个33的卷积核，在单通道图像上进行卷积，得到输出。

3D卷积：一个33*3的卷积核在立方体上进行卷积，得到输出。

传统的二维CNN中，卷积只应用在空间维度上，覆盖上一层的所有特征图，来计算二维判别特征图。HSI分类问题，希望能够同时捕获多个波段编码的光谱信息和空间信息。二维CNN不能处理光谱信息，而三维CNN可以在HSI数据中同时提取光谱和空间特征，但是代价是增加了计算复杂度。

代码实现

HybridSN 高光谱分类
S. K. Roy, G. Krishna, S. R. Dubey, B. B. Chaudhuri HybridSN: Exploring 3-D–2-D CNN Feature Hierarchy for Hyperspectral Image Classification, IEEE GRSL 2020

这篇论文构建了一个混合网络解决高光谱图像分类问题，首先用 3D卷积，然后使用 2D卷积，代码相对简单，下面是代码的解析。

取得数据，引入基本函数库

! wget http://www.ehu.eus/ccwintco/uploads/6/67/Indian_pines_corrected.mat
! wget http://www.ehu.eus/ccwintco/uploads/c/c4/Indian_pines_gt.mat
! pip install spectral

import numpy as np
import matplotlib.pyplot as plt
import scipy.io as sio
from sklearn.decomposition import PCA
from sklearn.model_selection import train_test_split
from sklearn.metrics import confusion_matrix, accuracy_score, classification_report, cohen_kappa_score
import spectral
import torch
import torchvision
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim

定义 HybridSN 类

模型的网络结构为如下图所示：

三维卷积部分：
conv1：（1, 30, 25, 25）， 8个 7x3x3 的卷积核 ==>（8, 24, 23, 23）
conv2：（8, 24, 23, 23）， 16个 5x3x3 的卷积核 ==>（16, 20, 21, 21）
conv3：（16, 20, 21, 21），32个 3x3x3 的卷积核 ==>（32, 18, 19, 19）
接下来要进行二维卷积，因此把前面的 32*18 reshape 一下，得到 （576, 19, 19）
二维卷积：（576, 19, 19） 64个 3x3 的卷积核，得到 （64, 17, 17）
接下来是一个 flatten 操作，变为 18496 维的向量，
接下来依次为256，128节点的全连接层，都使用比例为0.4的 Dropout，
最后输出为 16 个节点，是最终的分类类别数。
下面是 HybridSN 类的代码：

class_num = 16

class HybridSN(nn.Module):
  def __init__(self, num_classes=16):
    
    super(HybridSN, self).__init__()
    self.conv1 = nn.Sequential(nn.Conv3d(1, 8, (7, 3, 3)), nn.BatchNorm3d(8), nn.ReLU())
    self.conv2 = nn.Sequential(nn.Conv3d(8, 16, (5, 3, 3)), nn.BatchNorm3d(16), nn.ReLU())
    self.conv3 = nn.Sequential(nn.Conv3d(16, 32, (3, 3, 3)), nn.BatchNorm3d(32), nn.ReLU())
    self.conv3_2d = nn.Sequential(nn.Conv2d(576, 64, (3,3)), nn.BatchNorm2d(64), nn.ReLU())
    self.dense1 =  nn.Linear(18496,256)
    self.dense2 =  nn.Linear(256,128)
    self.out = nn.Linear(128, num_classes)
    self.drop = nn.Dropout(p=0.4)
    self.soft = nn.LogSoftmax(dim=1)
    self.relu = nn.ReLU()
    self.channel_attention_1 = ChannelAttention(576)
    self.spatial_attention_1 = SpatialAttention(kernel_size=7)
    self.channel_attention_2 = ChannelAttention(64)
    self.spatial_attention_2 = SpatialAttention(kernel_size=7)
  def forward(self, x):
    out = self.relu(self.conv1(x))
    out = self.relu(self.conv2(out))
    out = self.relu(self.conv3(out))
    # 进行二维卷积，因此把前面的 32*18 reshape 一下，得到 （576, 19, 19）
    out = out.view(-1, out.shape[1] * out.shape[2], out.shape[3], out.shape[4])
    out = self.channel_attention_1(out) * out
    out = self.spatial_attention_1(out) * out
    out = self.relu(self.conv3_2d(out))
    # flatten 操作，变为 18496 维的向量，
    out = out.view(out.size(0), -1)
    out = self.dense1(out)
    out = self.drop(out)
    out = self.dense2(out)
    out = self.drop(out)
    out = self.out(out)
    out = self.soft(out)
    return out

创建数据集

首先对高光谱数据实施PCA降维；然后创建 keras 方便处理的数据格式；然后随机抽取 10% 数据做为训练集，剩余的做为测试集。
首先定义基本函数：

# 对高光谱数据 X 应用 PCA 变换
def applyPCA(X, numComponents):
    newX = np.reshape(X, (-1, X.shape[2]))
    pca = PCA(n_components=numComponents, whiten=True)
    newX = pca.fit_transform(newX)
    newX = np.reshape(newX, (X.shape[0], X.shape[1], numComponents))
    return newX

# 对单个像素周围提取 patch 时，边缘像素就无法取了，因此，给这部分像素进行 padding 操作
def padWithZeros(X, margin=2):
    newX = np.zeros((X.shape[0] + 2 * margin, X.shape[1] + 2* margin, X.shape[2]))
    x_offset = margin
    y_offset = margin
    newX[x_offset:X.shape[0] + x_offset, y_offset:X.shape[1] + y_offset, :] = X
    return newX

# 在每个像素周围提取 patch ，然后创建成符合 keras 处理的格式
def createImageCubes(X, y, windowSize=5, removeZeroLabels = True):
    # 给 X 做 padding
    margin = int((windowSize - 1) / 2)
    zeroPaddedX = padWithZeros(X, margin=margin)
    # split patches
    patchesData = np.zeros((X.shape[0] * X.shape[1], windowSize, windowSize, X.shape[2]))
    patchesLabels = np.zeros((X.shape[0] * X.shape[1]))
    patchIndex = 0
    for r in range(margin, zeroPaddedX.shape[0] - margin):
        for c in range(margin, zeroPaddedX.shape[1] - margin):
            patch = zeroPaddedX[r - margin:r + margin + 1, c - margin:c + margin + 1]   
            patchesData[patchIndex, :, :, :] = patch
            patchesLabels[patchIndex] = y[r-margin, c-margin]
            patchIndex = patchIndex + 1
    if removeZeroLabels:
        patchesData = patchesData[patchesLabels>0,:,:,:]
        patchesLabels = patchesLabels[patchesLabels>0]
        patchesLabels -= 1
    return patchesData, patchesLabels

def splitTrainTestSet(X, y, testRatio, randomState=345):
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=testRatio, random_state=randomState, stratify=y)
    return X_train, X_test, y_train, y_test

下面读取并创建数据集：

# 地物类别
class_num = 16
X = sio.loadmat('Indian_pines_corrected.mat')['indian_pines_corrected']
y = sio.loadmat('Indian_pines_gt.mat')['indian_pines_gt']

# 用于测试样本的比例
test_ratio = 0.90
# 每个像素周围提取 patch 的尺寸
patch_size = 25
# 使用 PCA 降维，得到主成分的数量
pca_components = 30

print('Hyperspectral data shape: ', X.shape)
print('Label shape: ', y.shape)

print('\n... ... PCA tranformation ... ...')
X_pca = applyPCA(X, numComponents=pca_components)
print('Data shape after PCA: ', X_pca.shape)

print('\n... ... create data cubes ... ...')
X_pca, y = createImageCubes(X_pca, y, windowSize=patch_size)
print('Data cube X shape: ', X_pca.shape)
print('Data cube y shape: ', y.shape)

print('\n... ... create train & test data ... ...')
Xtrain, Xtest, ytrain, ytest = splitTrainTestSet(X_pca, y, test_ratio)
print('Xtrain shape: ', Xtrain.shape)
print('Xtest  shape: ', Xtest.shape)

# 改变 Xtrain, Ytrain 的形状，以符合 keras 的要求
Xtrain = Xtrain.reshape(-1, patch_size, patch_size, pca_components, 1)
Xtest  = Xtest.reshape(-1, patch_size, patch_size, pca_components, 1)
print('before transpose: Xtrain shape: ', Xtrain.shape) 
print('before transpose: Xtest  shape: ', Xtest.shape) 

# 为了适应 pytorch 结构，数据要做 transpose
Xtrain = Xtrain.transpose(0, 4, 3, 1, 2)
Xtest  = Xtest.transpose(0, 4, 3, 1, 2)
print('after transpose: Xtrain shape: ', Xtrain.shape) 
print('after transpose: Xtest  shape: ', Xtest.shape) 


""" Training dataset"""
class TrainDS(torch.utils.data.Dataset): 
    def __init__(self):
        self.len = Xtrain.shape[0]
        self.x_data = torch.FloatTensor(Xtrain)
        self.y_data = torch.LongTensor(ytrain)        
    def __getitem__(self, index):
        # 根据索引返回数据和对应的标签
        return self.x_data[index], self.y_data[index]
    def __len__(self): 
        # 返回文件数据的数目
        return self.len

""" Testing dataset"""
class TestDS(torch.utils.data.Dataset): 
    def __init__(self):
        self.len = Xtest.shape[0]
        self.x_data = torch.FloatTensor(Xtest)
        self.y_data = torch.LongTensor(ytest)
    def __getitem__(self, index):
        # 根据索引返回数据和对应的标签
        return self.x_data[index], self.y_data[index]
    def __len__(self): 
        # 返回文件数据的数目
        return self.len

# 创建 trainloader 和 testloader
trainset = TrainDS()
testset  = TestDS()
train_loader = torch.utils.data.DataLoader(dataset=trainset, batch_size=128, shuffle=True, num_workers=2)
test_loader  = torch.utils.data.DataLoader(dataset=testset,  batch_size=128, shuffle=False, num_workers=2)

开始训练

# 使用GPU训练，可以在菜单 "代码执行工具" -> "更改运行时类型" 里进行设置
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")

# 网络放到GPU上
net = HybridSN().to(device)
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(net.parameters(), lr=0.001)

# 开始训练
total_loss = 0
net.train() #启用训练模式
for epoch in range(100):
    for i, (inputs, labels) in enumerate(train_loader):
        inputs = inputs.to(device)
        labels = labels.to(device)
        # 优化器梯度归零
        optimizer.zero_grad()
        # 正向传播 +　反向传播 + 优化 
        outputs = net(inputs)
        loss = criterion(outputs, labels)
        loss.backward()
        optimizer.step()
        total_loss += loss.item()
    print('[Epoch: %d]   [loss avg: %.4f]   [current loss: %.4f]' %(epoch + 1, total_loss/(epoch+1), loss.item()))

print('Finished Training')

模型测试

count = 0
# 模型测试
net.eval()  #启用测试模式
for inputs, _ in test_loader:
    inputs = inputs.to(device)
    outputs = net(inputs)
    outputs = np.argmax(outputs.detach().cpu().numpy(), axis=1)
    if count == 0:
        y_pred_test =  outputs
        count = 1
    else:
        y_pred_test = np.concatenate( (y_pred_test, outputs) )

# 生成分类报告
classification = classification_report(ytest, y_pred_test, digits=4)
print(classification)

备用函数

下面是用于计算各个类准确率，显示结果的备用函数，以供参考

from operator import truediv

def AA_andEachClassAccuracy(confusion_matrix):
    counter = confusion_matrix.shape[0]
    list_diag = np.diag(confusion_matrix)
    list_raw_sum = np.sum(confusion_matrix, axis=1)
    each_acc = np.nan_to_num(truediv(list_diag, list_raw_sum))
    average_acc = np.mean(each_acc)
    return each_acc, average_acc


def reports (test_loader, y_test, name):
    count = 0
    # 模型测试
    for inputs, _ in test_loader:
        inputs = inputs.to(device)
        outputs = net(inputs)
        outputs = np.argmax(outputs.detach().cpu().numpy(), axis=1)
        if count == 0:
            y_pred =  outputs
            count = 1
        else:
            y_pred = np.concatenate( (y_pred, outputs) )

    if name == 'IP':
        target_names = ['Alfalfa', 'Corn-notill', 'Corn-mintill', 'Corn'
                        ,'Grass-pasture', 'Grass-trees', 'Grass-pasture-mowed', 
                        'Hay-windrowed', 'Oats', 'Soybean-notill', 'Soybean-mintill',
                        'Soybean-clean', 'Wheat', 'Woods', 'Buildings-Grass-Trees-Drives',
                        'Stone-Steel-Towers']
    elif name == 'SA':
        target_names = ['Brocoli_green_weeds_1','Brocoli_green_weeds_2','Fallow','Fallow_rough_plow','Fallow_smooth',
                        'Stubble','Celery','Grapes_untrained','Soil_vinyard_develop','Corn_senesced_green_weeds',
                        'Lettuce_romaine_4wk','Lettuce_romaine_5wk','Lettuce_romaine_6wk','Lettuce_romaine_7wk',
                        'Vinyard_untrained','Vinyard_vertical_trellis']
    elif name == 'PU':
        target_names = ['Asphalt','Meadows','Gravel','Trees', 'Painted metal sheets','Bare Soil','Bitumen',
                        'Self-Blocking Bricks','Shadows']
    
    classification = classification_report(y_test, y_pred, target_names=target_names)
    oa = accuracy_score(y_test, y_pred)
    confusion = confusion_matrix(y_test, y_pred)
    each_acc, aa = AA_andEachClassAccuracy(confusion)
    kappa = cohen_kappa_score(y_test, y_pred)
    
    return classification, confusion, oa*100, each_acc*100, aa*100, kappa*100

检测结果写在文件里

classification, confusion, oa, each_acc, aa, kappa = reports(test_loader, ytest, 'IP')
classification = str(classification)
confusion = str(confusion)
file_name = "classification_report.txt"

with open(file_name, 'w') as x_file:
    x_file.write('\n')
    x_file.write('{} Kappa accuracy (%)'.format(kappa))
    x_file.write('\n')
    x_file.write('{} Overall accuracy (%)'.format(oa))
    x_file.write('\n')
    x_file.write('{} Average accuracy (%)'.format(aa))
    x_file.write('\n')
    x_file.write('\n')
    x_file.write('{}'.format(classification))
    x_file.write('\n')
    x_file.write('{}'.format(confusion))

下面代码用于显示分类结果：

# load the original image
X = sio.loadmat('Indian_pines_corrected.mat')['indian_pines_corrected']
y = sio.loadmat('Indian_pines_gt.mat')['indian_pines_gt']

height = y.shape[0]
width = y.shape[1]

X = applyPCA(X, numComponents= pca_components)
X = padWithZeros(X, patch_size//2)

# 逐像素预测类别
outputs = np.zeros((height,width))
for i in range(height):
    for j in range(width):
        if int(y[i,j]) == 0:
            continue
        else :
            image_patch = X[i:i+patch_size, j:j+patch_size, :]
            image_patch = image_patch.reshape(1,image_patch.shape[0],image_patch.shape[1], image_patch.shape[2], 1)
            X_test_image = torch.FloatTensor(image_patch.transpose(0, 4, 3, 1, 2)).to(device)                                   
            prediction = net(X_test_image)
            prediction = np.argmax(prediction.detach().cpu().numpy(), axis=1)
            outputs[i][j] = prediction+1
    if i % 20 == 0:
        print('... ... row ', i, ' handling ... ...')

predict_image = spectral.imshow(classes = outputs.astype(int),figsize =(5,5))

思考问题二：分类结果不同的原因

训练网络，然后多测试几次，会发现每次分类的结果都不一样，请思考为什么？
未修改前2次训练结果图


这是因为原始的代码没有注意到pytorch模型网络的训练模式和测试模式，由于训练的时候Dropout和BN层起作用，每个batch BN层的参数不一样，dropout在训练时随机失效点具有随机性，所以训练和测试要区分开来。在本例中因为网络的全连接层中为了避免过拟合使用nn.Dropout(p=0.4)，而原始代码在测试的时候没有启用测试模式，即使用的是训练模式，所以每次结果都是不一样的。因此在训练和测试的时候分别启用相应模式，即 net.train() 和 net.eval() (这里net是我们构建的HybirdSN)，就可以保证测试的结果是一样的了。前面的结果是修改后的。

思考问题三：注意力机制

思考问题，如果想要进一步提升高光谱图像的分类性能，可以如何使用注意力机制？
可以从图像空间、通道和光谱波长三个方向使用注意力机制来处理特征，进而提升高光谱图像的分类性能。

# 通道注意力机制
class ChannelAttention(nn.Module):
    def __init__(self, in_planes, ratio=16):
        super(ChannelAttention, self).__init__()
        self.avg_pool = nn.AdaptiveAvgPool2d(1)
        self.max_pool = nn.AdaptiveMaxPool2d(1)
 
        self.fc1   = nn.Conv2d(in_planes, in_planes // 16, 1, bias=False)
        self.relu1 = nn.ReLU()
        self.fc2   = nn.Conv2d(in_planes // 16, in_planes, 1, bias=False)
 
        self.sigmoid = nn.Sigmoid()
 
    def forward(self, x):
        avg_out = self.fc2(self.relu1(self.fc1(self.avg_pool(x))))
        max_out = self.fc2(self.relu1(self.fc1(self.max_pool(x))))
        out = avg_out + max_out
        return self.sigmoid(out)
# 空间注意力机制
class SpatialAttention(nn.Module):
    def __init__(self, kernel_size=7):
        super(SpatialAttention, self).__init__()
 
        assert kernel_size in (3, 7), 'kernel size must be 3 or 7'
        padding = 3 if kernel_size == 7 else 1
 
        self.conv1 = nn.Conv2d(2, 1, kernel_size, padding=padding, bias=False)
        self.sigmoid = nn.Sigmoid()
 
    def forward(self, x):
        avg_out = torch.mean(x, dim=1, keepdim=True)
        max_out, _ = torch.max(x, dim=1, keepdim=True)
        x = torch.cat([avg_out, max_out], dim=1)
        x = self.conv1(x)
        return self.sigmoid(x)

可以看到准确率从0.9551提高到0.9778，是有明显提升的。

仔细观察可以看出两次的图片左上方有明显区别，在细节上效果更好。