UCM数据集21类分类

import cv2
import numpy as np
from sklearn.metrics import classification_report
from sklearn.model_selection import train_test_split
import matplotlib.pyplot as plt
from torchvision.datasets import ImageFolder

dataset = ImageFolder('E:\\pycharm\\data\\Images\\')
a = cv2.imread('E:/pycharm/data/Images/airplane00.tif')
# print(a.shape)
dataset = dataset.imgs
print(dataset)

data_path = []

data_label = []
for i in range(dataset.__len__()):
    # print(dataset[i][0])
    # print(dataset[i][1])
    data_path.append(dataset[i][0])
    data_label.append(dataset[i][1])
# print(data_path,'\n',data_label)

data_arr=[]
for i in range(len(data_path)):
    # print(data_path[i])
    p = cv2.imread(data_path[i])
    p = cv2.resize(p,(224,224))
    # print(p.shape)
    # print(p)
    # data_arr[i]=np.array(p).resize(224*224*3)
    data_arr.append(np.array(p).flatten())

print(type(data_arr))
data_arr  = np.array(data_arr)
data_label = np.array(data_label)
#
# print(data_arr)


X_train, X_test, y_train, y_test = train_test_split(data_arr, data_label, test_size=0.1, random_state=200)

from sklearn import svm
predictor = svm.SVC(gamma='scale', C=1.0,max_iter = 1000)
# 进行训练
predictor.fit(X_train, y_train)
predictions_labels = predictor.predict(X_test)
print(classification_report(y_test, predictions_labels,digits=4))