环境：ubuntu14.04+python3.5+anaconda+opencv3.1 本帖基于论文：Low-effort place recognition with WiFi fingerprints using Deep Learning
室内定位有很多种方式，利用WiFi指纹就是是其中的一种。在室内，可以通过WiFi信号强度来确定移动设备的大致位置，参看：https://www.zhihu.com/question/20593603
使用WiFi指纹定位的简要流程
首先采集WiFi信号，这并不需要什么专业的设备，几台手机即可。Android手机上有很多检测WiFi的App，如Sensor Log。
把室内划分成网格块(对应位置)，站在每个块内分别使用Sensor Log检测WiFi信号，数据越多越好。如下：
location1 : WiFi{"BSSID":"11:11:11:11:11:11",...."level":-33,"....} # 所在位置对应的AP,RSSI信号强度等信息
location2 : WiFi{"BSSID":"11:11:11:11:11:11",...."level":-27,"....}
location2 : WiFi{"BSSID":"22:22:22:22:22:22",...."level":-80,"....}
location3 : WiFi{"BSSID":"22:22:22:22:22:22",...."level":-54,"....}
...
无线信号强度是负值，范围一般在0<->-90dbm。值越大信号越强，-50dbm强于-70dbm，
数据采集完成之后，对数据进行预处理，制作成WiFi指纹数据库，参考下面的UJIIndoorLoc数据集。
开发分类模型(本帖关注点)。
最后，用户上传所在位置的wifi信息，分类模型返回预测的位置。
TensorFlow练习: 基于WiFi指纹的室内定位
使用的数据集：https://archive.ics.uci.edu/ml/datasets/UJIIndoorLoc 下载数据集：
wget https://archive.ics.uci.edu/ml/machine-learning-databases/00310/UJIndoorLoc.zip
unzip UJIndoorLoc.zip
代码
import tensorflow as tf
import numpy as np
import cv2
inception_model = 'tensorflow_inception_graph.pb'
# 加载inception模型
graph = tf.Graph()
sess = tf.InteractiveSession(graph=graph)
X = tf.placeholder(np.float32, name='input')
with tf.gfile.FastGFile(inception_model, 'rb') as f:
graph_def = tf.GraphDef()
graph_def.ParseFromString(f.read())
imagenet_mean = 117.0
preprocessed = tf.expand_dims(X-imagenet_mean, 0)
tf.import_graph_def(graph_def, {'input':preprocessed})
layers = [op.name for op in graph.get_operations() if op.type=='Conv2D' and 'import/' in op.name]
feature_nums = [int(graph.get_tensor_by_name(name+':0').get_shape()[-1]) for name in layers]
print('layers:', len(layers)) # 59
print('feature:', sum(feature_nums)) # 7548
# deep dream
def deep_dream(obj, img_noise=np.random.uniform(size=(224,224,3)) + 100.0, iter_n=10, step=1.5, octave_n=4, octave_scale=1.4):
score = tf.reduce_mean(obj)
gradi = tf.gradients(score, X)[0]
img = img_noise
octaves = []
def tffunc(*argtypes):
placeholders = list(map(tf.placeholder, argtypes))
def wrap(f):
out = f(*placeholders)
def wrapper(*args, **kw):
return out.eval(dict(zip(placeholders, args)), session=kw.get('session'))
return wrapper
return wrap
def resize(img, size):
img = tf.expand_dims(img, 0)
return tf.image.resize_bilinear(img, size)[0,:,:,:]
resize = tffunc(np.float32, np.int32)(resize)
for _ in range(octave_n-1):
hw = img.shape[:2]
lo = resize(img, np.int32(np.float32(hw)/octave_scale))
hi = img-resize(lo, hw)
img = lo
octaves.append(hi)
def calc_grad_tiled(img, t_grad, tile_size=512):
sz = tile_size
h, w = img.shape[:2]
sx, sy = np.random.randint(sz, size=2)
img_shift = np.roll(np.roll(img, sx, 1), sy, 0)
grad = np.zeros_like(img)
for y in range(0, max(h-sz//2, sz),sz):
for x in range(0, max(w-sz//2, sz),sz):
sub = img_shift[y:y+sz,x:x+sz]
g = sess.run(t_grad, {X:sub})
grad[y:y+sz,x:x+sz] = g
return np.roll(np.roll(grad, -sx, 1), -sy, 0)
for octave in range(octave_n):
if octave>0:
hi = octaves[-octave]
img = resize(img, hi.shape[:2])+hi
for _ in range(iter_n):
g = calc_grad_tiled(img, gradi)
img += g*(step / (np.abs(g).mean()+1e-7))
# 保存图像
output_file = 'output' + str(octave+1) + '.jpg'
cv2.imwrite(output_file, img)
print(output_file)
# 加载输入图像
input_img = cv2.imread('input.jpg')
input_img = np.float32(input_img)
# 选择层
layer = 'mixed4c'
deep_dream(tf.square(graph.get_tensor_by_name("import/%s:0"%layer)), input_img)
训练过程：
Epoch: 0 Loss: 0.813012358957
Epoch: 1 Loss: 0.745993956116
Epoch: 2 Loss: 0.730452445426
Epoch: 3 Loss: 0.722375573073
Epoch: 4 Loss: 0.71705802879
转自：ttp://blog.topspeedsnail.com/archives/10468
