我需要一种算法,可以确定两个图像是否"相似"并识别相似的颜色,亮度,形状等模式.我可能需要一些关于人类大脑用来"分类"图像的参数的指针...
我已经看过基于hausdorff的匹配,但这似乎主要是为了匹配变换对象和形状模式.
通过使用小波变换将图像分解为签名,我做了类似的事情.
我的方法是从每个变换的通道中选择最重要的n个系数,并记录它们的位置.这是通过根据abs(功率)对(功率,位置)元组列表进行排序来完成的.类似的图像将具有相似之处,因为它们在相同的位置具有显着的系数.
我发现最好将图像转换为YUV格式,这有效地允许您在形状(Y通道)和颜色(UV通道)中加权相似性.
你可以在mactorii中找到我对上面的实现,遗憾的是我没有像我应该的那样工作:-)
另一种方法,我的一些朋友已经使用了令人惊讶的好结果,只是简单地调整你的图像,4x4像素和存储,这是你的签名.可以通过使用相应的像素计算2个图像之间的曼哈顿距离来对2个图像的相似程度进行评分.我没有详细说明他们如何执行调整大小,因此您可能必须使用可用于该任务的各种算法来找到合适的算法.
pHash可能会让你感兴趣.
感知哈希 音频,视频或图像文件的指纹,其在数学上基于其中包含的音频或视觉内容.与依赖于输入的微小变化导致输出剧烈变化的雪崩效应的加密散列函数不同,如果输入在视觉上或听觉上相似,则感知散列彼此"接近".
我使用SIFT重新检测不同图像中的同一个对象.它真的很强大但相当复杂,可能有点矫枉过正.如果图像应该非常相似,那么基于两个图像之间的差异的一些简单参数可以告诉你很多.一些指示:
归一化图像,即通过计算两者的平均亮度使两个图像的平均亮度相同,并根据比例缩小最亮度(以避免在最高级别剪裁)),尤其是如果您对形状比对形状更感兴趣颜色.
每个通道的归一化图像的色差之和.
找到图像中的边缘并测量两个图像中边缘像素之间的距离.(用于形状)
将图像划分为一组离散区域,并比较每个区域的平均颜色.
在一个(或一组)级别对图像进行阈值处理,并计算产生的黑白图像不同的像素数.
您可以使用感知图像差异
它是一个命令行实用程序,使用感知度量标准比较两个图像.也就是说,它使用人类视觉系统的计算模型来确定两个图像是否在视觉上不同,因此忽略了像素的微小变化.此外,它大大减少了随机数生成,操作系统或机器架构差异造成的误报数量.
我的实验室也需要解决此问题，因此我们使用了Tensorflow。这是用于可视化图像相似性的完整应用程序实现。
有关将图像矢量化以进行相似度计算的教程，请查看此页面。这是Python（同样，请参见该帖子以获取完整的工作流程）：
from __future__ import absolute_import, division, print_function
"""
This is a modification of the classify_images.py
script in Tensorflow. The original script produces
string labels for input images (e.g. you input a picture
of a cat and the script returns the string "cat"); this
modification reads in a directory of images and
generates a vector representation of the image using
the penultimate layer of neural network weights.
Usage: python classify_images.py "../image_dir/*.jpg"
"""
# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Simple image classification with Inception.
Run image classification with Inception trained on ImageNet 2012 Challenge data
set.
This program creates a graph from a saved GraphDef protocol buffer,
and runs inference on an input JPEG image. It outputs human readable
strings of the top 5 predictions along with their probabilities.
Change the --image_file argument to any jpg image to compute a
classification of that image.
Please see the tutorial and website for a detailed description of how
to use this script to perform image recognition.
https://tensorflow.org/tutorials/image_recognition/
"""
import os.path
import re
import sys
import tarfile
import glob
import json
import psutil
from collections import defaultdict
import numpy as np
from six.moves import urllib
import tensorflow as tf
FLAGS = tf.app.flags.FLAGS
# classify_image_graph_def.pb:
# Binary representation of the GraphDef protocol buffer.
# imagenet_synset_to_human_label_map.txt:
# Map from synset ID to a human readable string.
# imagenet_2012_challenge_label_map_proto.pbtxt:
# Text representation of a protocol buffer mapping a label to synset ID.
tf.app.flags.DEFINE_string(
'model_dir', '/tmp/imagenet',
"""Path to classify_image_graph_def.pb, """
"""imagenet_synset_to_human_label_map.txt, and """
"""imagenet_2012_challenge_label_map_proto.pbtxt.""")
tf.app.flags.DEFINE_string('image_file', '',
"""Absolute path to image file.""")
tf.app.flags.DEFINE_integer('num_top_predictions', 5,
"""Display this many predictions.""")
# pylint: disable=line-too-long
DATA_URL = 'http://download.tensorflow.org/models/image/imagenet/inception-2015-12-05.tgz'
# pylint: enable=line-too-long
class NodeLookup(object):
"""Converts integer node ID's to human readable labels."""
def __init__(self,
label_lookup_path=None,
uid_lookup_path=None):
if not label_lookup_path:
label_lookup_path = os.path.join(
FLAGS.model_dir, 'imagenet_2012_challenge_label_map_proto.pbtxt')
if not uid_lookup_path:
uid_lookup_path = os.path.join(
FLAGS.model_dir, 'imagenet_synset_to_human_label_map.txt')
self.node_lookup = self.load(label_lookup_path, uid_lookup_path)
def load(self, label_lookup_path, uid_lookup_path):
"""Loads a human readable English name for each softmax node.
Args:
label_lookup_path: string UID to integer node ID.
uid_lookup_path: string UID to human-readable string.
Returns:
dict from integer node ID to human-readable string.
"""
if not tf.gfile.Exists(uid_lookup_path):
tf.logging.fatal('File does not exist %s', uid_lookup_path)
if not tf.gfile.Exists(label_lookup_path):
tf.logging.fatal('File does not exist %s', label_lookup_path)
# Loads mapping from string UID to human-readable string
proto_as_ascii_lines = tf.gfile.GFile(uid_lookup_path).readlines()
uid_to_human = {}
p = re.compile(r'[n\d]*[ \S,]*')
for line in proto_as_ascii_lines:
parsed_items = p.findall(line)
uid = parsed_items[0]
human_string = parsed_items[2]
uid_to_human[uid] = human_string
# Loads mapping from string UID to integer node ID.
node_id_to_uid = {}
proto_as_ascii = tf.gfile.GFile(label_lookup_path).readlines()
for line in proto_as_ascii:
if line.startswith(' target_class:'):
target_class = int(line.split(': ')[1])
if line.startswith(' target_class_string:'):
target_class_string = line.split(': ')[1]
node_id_to_uid[target_class] = target_class_string[1:-2]
# Loads the final mapping of integer node ID to human-readable string
node_id_to_name = {}
for key, val in node_id_to_uid.items():
if val not in uid_to_human:
tf.logging.fatal('Failed to locate: %s', val)
name = uid_to_human[val]
node_id_to_name[key] = name
return node_id_to_name
def id_to_string(self, node_id):
if node_id not in self.node_lookup:
return ''
return self.node_lookup[node_id]
def create_graph():
"""Creates a graph from saved GraphDef file and returns a saver."""
# Creates graph from saved graph_def.pb.
with tf.gfile.FastGFile(os.path.join(
FLAGS.model_dir, 'classify_image_graph_def.pb'), 'rb') as f:
graph_def = tf.GraphDef()
graph_def.ParseFromString(f.read())
_ = tf.import_graph_def(graph_def, name='')
def run_inference_on_images(image_list, output_dir):
"""Runs inference on an image list.
Args:
image_list: a list of images.
output_dir: the directory in which image vectors will be saved
Returns:
image_to_labels: a dictionary with image file keys and predicted
text label values
"""
image_to_labels = defaultdict(list)
create_graph()
with tf.Session() as sess:
# Some useful tensors:
# 'softmax:0': A tensor containing the normalized prediction across
# 1000 labels.
# 'pool_3:0': A tensor containing the next-to-last layer containing 2048
# float description of the image.
# 'DecodeJpeg/contents:0': A tensor containing a string providing JPEG
# encoding of the image.
# Runs the softmax tensor by feeding the image_data as input to the graph.
softmax_tensor = sess.graph.get_tensor_by_name('softmax:0')
for image_index, image in enumerate(image_list):
try:
print("parsing", image_index, image, "\n")
if not tf.gfile.Exists(image):
tf.logging.fatal('File does not exist %s', image)
with tf.gfile.FastGFile(image, 'rb') as f:
image_data = f.read()
predictions = sess.run(softmax_tensor,
{'DecodeJpeg/contents:0': image_data})
predictions = np.squeeze(predictions)
###
# Get penultimate layer weights
###
feature_tensor = sess.graph.get_tensor_by_name('pool_3:0')
feature_set = sess.run(feature_tensor,
{'DecodeJpeg/contents:0': image_data})
feature_vector = np.squeeze(feature_set)
outfile_name = os.path.basename(image) + ".npz"
out_path = os.path.join(output_dir, outfile_name)
np.savetxt(out_path, feature_vector, delimiter=',')
# Creates node ID --> English string lookup.
node_lookup = NodeLookup()
top_k = predictions.argsort()[-FLAGS.num_top_predictions:][::-1]
for node_id in top_k:
human_string = node_lookup.id_to_string(node_id)
score = predictions[node_id]
print("results for", image)
print('%s (score = %.5f)' % (human_string, score))
print("\n")
image_to_labels[image].append(
{
"labels": human_string,
"score": str(score)
}
)
# close the open file handlers
proc = psutil.Process()
open_files = proc.open_files()
for open_file in open_files:
file_handler = getattr(open_file, "fd")
os.close(file_handler)
except:
print('could not process image index',image_index,'image', image)
return image_to_labels
def maybe_download_and_extract():
"""Download and extract model tar file."""
dest_directory = FLAGS.model_dir
if not os.path.exists(dest_directory):
os.makedirs(dest_directory)
filename = DATA_URL.split('/')[-1]
filepath = os.path.join(dest_directory, filename)
if not os.path.exists(filepath):
def _progress(count, block_size, total_size):
sys.stdout.write('\r>> Downloading %s %.1f%%' % (
filename, float(count * block_size) / float(total_size) * 100.0))
sys.stdout.flush()
filepath, _ = urllib.request.urlretrieve(DATA_URL, filepath, _progress)
print()
statinfo = os.stat(filepath)
print('Succesfully downloaded', filename, statinfo.st_size, 'bytes.')
tarfile.open(filepath, 'r:gz').extractall(dest_directory)
def main(_):
maybe_download_and_extract()
if len(sys.argv) < 2:
print("please provide a glob path to one or more images, e.g.")
print("python classify_image_modified.py '../cats/*.jpg'")
sys.exit()
else:
output_dir = "image_vectors"
if not os.path.exists(output_dir):
os.makedirs(output_dir)
images = glob.glob(sys.argv[1])
image_to_labels = run_inference_on_images(images, output_dir)
with open("image_to_labels.json", "w") as img_to_labels_out:
json.dump(image_to_labels, img_to_labels_out)
print("all done")
if __name__ == '__main__':
tf.app.run()
京公网安备 11010802040832号 | 京ICP备19059560号-6 