Skip to main content
 首页 » 编程设计

python之PCA 后如何从 Python 中的第一个组件重建图像

2024年11月01日8fff_TT

如何在 PCA 之后用 Python 中的第一个组件重建图像?

我的尝试:

from sklearn.datasets import load_sample_image 
from sklearn.feature_extraction import image 
from sklearn.decomposition import PCA 
# Create patches of size 25 by 25 and create a matrix from all patches 
patches = image.extract_patches_2d(grayscale_image, (25, 25), random_state = 42) 
#reshape patches 
patches_reshaped = patches.reshape(patches.shape[0],-1) 
# PCA 
pca = PCA(n_components = 3,random_state = 42) 
pca.fit(patches_reshaped) 
first_component = pca.components_[0] #first component 
# attempt to reconstruct image from first component 
plt.imshow(pca.components_[0].reshape(25, 25),"gray") 

请您参考如下方法:

您已将 n_components = 3 传递给 PCA,这意味着您将拥有三个主成分。所以当你这样做的时候,

projected = pca.fit_transform(patches_reshaped.data) 

您将在三个 3 个主轴上投影数据,这意味着您的输出将具有形状 (patches.shape[0], 3)。

现在要使用第一个主成分进行重建,您需要做的是将数据投影到这个主轴上,并对原始域进行逆变换。为此,首先获取第一个主成分:

# First get your first component 
first_component = pca.components_[0] 
# Make sure your first component is a row vector 
first_component = first_component.reshape(1,-1)  

然后,逆变换就是projected_data * principal_components。有关更多详细信息,您可以查看文档 here和源代码 here .

# get the first projection  
first_proj = projected[:,0] 
# Make sure your first component is a column vector 
first_proj = first_proj.reshape(-1,1) 
# do inverse transform (No you have to add the mean as thse algorithm  
# works on zero mean data)  
recon_using_first_comp = np.dot(proj, first_component) + pca.mean_ 

然后重建补丁以获得最终图像

final_img = image.reconstruct_from_patches_2d(recon_using_first_comp.reshape(-1,25,25), grayscale_image.shape)