def gen_train_val_data(sample_dir, train_rate, num_classes, condition_classes):
train_data, train_label, valid_data, valid_label, train_n, valid_n, note_label = load_image(sample_dir, 1.0).gen_train_valid()
train_data_trip = create_pairs(train_data, num_classes, condition_classes)
train_num = int(train_n*train_rate)
train_data_trip = train_data_trip[0:train_num]
val_data_trip = train_data_trip[-train_num:-1]
test_data_trip = val_data_trip
return train_data_trip, val_data_trip, test_data_trip
def main():
print(df)
for index, row in df.iterrows():
img = load_image(row['image_name'] + '.tif')
print(type(img)) #np.ndarray
print(img.shape)
print(row['image_name'])
r, g, b, nir = img[:, :, 0], img[:, :, 1], img[:, :, 2], img[:, :, 3]
fig = plt.figure()
fig.set_size_inches(12, 4)
for i, (x, c) in enumerate(
((r, 'r'), (g, 'g'), (b, 'b'), (nir, 'near-ir'))):
a = fig.add_subplot(1, 4, i + 1)
a.set_title(c)
plt.imshow(x)
plt.gray()
plt.show()
if index > 1: break
Y_train = df['tags'].values
"""
Created on 2017 10.17
@author: liupeng
"""
import sys
import tensorflow as tf
import numpy as np
import os
import cv2
from skimage import exposure
from data_load import load_image
# 图片在二级目录
file_path = 'gender'
img_path, _, _, _, _, _, _ = load_image(file_path, 1.0).gen_train_valid()
print(img_path)
def add_img_padding(img):
h, w, _ = img.shape
width = np.max([h, w])
# 按照长宽中大的初始化一个正方形
img_padding = np.zeros([width, width, 3])
# 找出padding的中间位置
h1 = int(width / 2 - h / 2)
h2 = int(width / 2 + h / 2)
w1 = int(width / 2 - w / 2)
w2 = int(width / 2 + w / 2)
# 进行padding， img为什么也要进行扣取？ 原因在于除以2取整会造成尺寸不一致的情况。
img_padding[h1:h2, w1:w2, :] = img[0:(h2 - h1), 0:(w2 - w1), :]
# coding: utf-8
import pandas as pd
import numpy as np
import data_load
import feature_extract
from sklearn.model_selection import train_test_split
from sklearn.pipeline import Pipeline
from sklearn.linear_model import LinearRegression
from sklearn.preprocessing import StandardScaler
from sklearn.decomposition import KernelPCA
from sklearn
if __name__=="__main__":
train_df = data_load.load_train()
train_df = train_df.sample(40)
X = [ data_load.load_image(row.image_name + ".jpg") for i, row in train_df.iterrows()]
hog_X = np.array([ feature_extract.hog_cnl(x) for x in X ])
y = np.array([ row.tags for i, row in train_df.iterrows()])
print(hog_X.shape)
print(y.shape)
X_dev, X_val, y_dev, y_val = train_test_split(hog_X, y, test_size=0.2)
pipe = Pipeline([('sc', StandardScaler()),
('pca', KernelPCA(n_components=200, kernel="linear")),
('clf', LinearRegression())])