深海游弋的鱼 – 第 10 页

macOS High Sierra下使用brew安装支持x265的ffmpeg以及转码

使用

$ brew install ffmpeg

1	$ brew install ffmpeg

安装ffmpeg默认是不支持x265的

使用

$ brew reinstall ffmpeg --with-x265

1	$ brew reinstall ffmpeg --with-x265

重新安装即可。

一般直接的H265数据流(比如从摄像头抓取的H265裸数据流保存的文件），是没办法在MacOS上直接播放的，必须进行转码。简单的转码命令如下：

$ ffmpeg -i input.h265 -c:v libx265 output.mp4

1	$ ffmpeg -i input.h265 -c:v libx265 output.mp4

参考链接

关闭Safari英文矫正功能

在macOS High Sierra系统上使用Safari进行搜索的时候，总是莫名其妙的被修改搜索词，导致搜索结果驴唇不对马嘴。搜索了一下，这个锅应该是输入法导致的，最可恶的是不能拒绝输入法的推荐词汇，可以使用如下方式关闭输入法的自动英文矫正功能。

继续阅读关闭Safari英文矫正功能

macOS Sierra升级到macOS High Sierra后执行cc,git命令报错

macOS Sierra升级到macOS High Sierra后执行cc,git等命令报错，错误信息如下：

xcrun: error: invalid active developer path (/Library/Developer/CommandLineTools), missing xcrun at: /Library/Developer/CommandLineTools/usr/bin/xcrun

1	xcrun: error: invalid active developer path (/Library/Developer/CommandLineTools), missing xcrun at: /Library/Developer/CommandLineTools/usr/bin/xcrun

解决方法如下：

$ xcode-select --install

1	$ xcode-select --install

参考链接

mac更新系统后Git不能用，提示missing xcrun at

macOS High Sierra (10.13)制作Windows 7安装U盘

Open a Terminal (under Utilities)
Convert the ISO to UDRW format hdiutil convert -format UDRW -o destination_file.img source_file.iso
Run diskutil list and determine the device node assigned to your flash media (e.g., /dev/disk2)
Run diskutil unmountDisk /dev/diskN (replace N with the disk number from the last command; in the previous example, N would be 2)
Execute sudo dd if=/path/to/destination_file.img.dmg of=/dev/diskN bs=1m (replace /path/to/destination_file.img.dmg with the path where the image file is located; for example, ./win7.img.dmg)
- Using /dev/rdisk instead of /dev/disk will faster, 20x
- If you see the error dd: Invalid number '1m', you are using GNU dd. Use the same command but replace bs=1m with bs=1M
- If you see the error dd: /dev/diskN: Resource busy, make sure the disk is not in use. Start the 'Disk Utility.app' and unmount (don't eject) the drive
Run diskutil eject /dev/diskN, and remove your flash media when the command completes (this can take a few hours on slower drives)

继续阅读macOS High Sierra (10.13)制作Windows 7安装U盘

获取当前Python中site-packages的具体存放路径

很多时候，我们系统上安装了好几个版本的Python, 此时，我们往往没办法确定通过pip安装的包会存放到那个目录下的site-packages中，可以通过如下代码获取：

$ python -c "from distutils.sysconfig import get_python_lib; print get_python_lib()"

1	$ python -c "from distutils.sysconfig import get_python_lib; print get_python_lib()"

在Ubuntu 16.04 LTS系统上，这个输出是存在问题的，执行命令后输出的目录是：

/usr/lib/python2.7/dist-packages

1	/usr/lib/python2.7/dist-packages

实际上，通过pip命令安装的目录有很大一部分被安装到了

/usr/local/lib/python2.7/dist-packages

1	/usr/local/lib/python2.7/dist-packages

目录下。

macOS Sierra (10.12.4)下使用Android Studio打开Android Virtual Device Manager报告“/dev/kvm is not found”错误

Android Virtual Device Manager突然出现了/dev/kvm is not found这个错误，我猜测大概Hardware_Accelerated_Execution_Manager丢失了某些文件，或者没安装好HAXM。

继续阅读macOS Sierra (10.12.4)下使用Android Studio打开Android Virtual Device Manager报告“/dev/kvm is not found”错误

macOS Sierra (10.12.4)下Caffe执行Python代码报告错误“Mean shape incompatible with input shape”

在执行macOS Sierra (10.12.4)下Caffe通过Python接口加载binaryproto格式的均值文件的时候，最后报告错误：

Traceback (most recent call last):
  File "analysis_memnet.py", line 29, in <module>
    detector = caffe.Detector(model_def, pretrained_model, mean=means)
  File "/Users/Source/caffe/distribute/python/caffe/detector.py", line 46, in __init__
    self.transformer.set_mean(in_, mean)
  File "/Users/Source/caffe/distribute/python/caffe/io.py", line 259, in set_mean
    raise ValueError('Mean shape incompatible with input shape.')
ValueError: Mean shape incompatible with input shape.

Traceback (most recent call last):

File "analysis_memnet.py", line 29, in <module>

detector = caffe.Detector(model_def, pretrained_model, mean=means)

File "/Users/Source/caffe/distribute/python/caffe/detector.py", line 46, in __init__

self.transformer.set_mean(in_, mean)

File "/Users/Source/caffe/distribute/python/caffe/io.py", line 259, in set_mean

raise ValueError('Mean shape incompatible with input shape.')

ValueError: Mean shape incompatible with input shape.

这个错误发生的原因是由于memnet提供的均值文件是256*256的，但是提供的配置文件却是227*227的，导致在io.py里面的代码在进行判断的时候发生异常。调整源代码中的python/caffe/io.py里面的代码：

    def set_mean(self, in_, mean):
        """
        Set the mean to subtract for centering the data.

        Parameters
        ----------
        in_ : which input to assign this mean.
        mean : mean ndarray (input dimensional or broadcastable)
        """
        self.__check_input(in_)
        ms = mean.shape
        if mean.ndim == 1:
            # broadcast channels
            if ms[0] != self.inputs[in_][1]:
                raise ValueError('Mean channels incompatible with input.')
            mean = mean[:, np.newaxis, np.newaxis]
        else:
            # elementwise mean
            if len(ms) == 2:
                ms = (1,) + ms
            if len(ms) != 3:
                raise ValueError('Mean shape invalid')
            if ms != self.inputs[in_][1:]:
                raise ValueError('Mean shape incompatible with input shape.')
        self.mean[in_] = mean

def set_mean(self, in_, mean):

"""

Set the mean to subtract for centering the data.

Parameters

----------

in_ : which input to assign this mean.

mean : mean ndarray (input dimensional or broadcastable)

"""

self.__check_input(in_)

ms = mean.shape

if mean.ndim == 1:

# broadcast channels

if ms[0] != self.inputs[in_][1]:

raise ValueError('Mean channels incompatible with input.')

mean = mean[:, np.newaxis, np.newaxis]

else:

# elementwise mean

if len(ms) == 2:

ms = (1,) + ms

if len(ms) != 3:

raise ValueError('Mean shape invalid')

if ms != self.inputs[in_][1:]:

raise ValueError('Mean shape incompatible with input shape.')

self.mean[in_] = mean

调整为：

    def set_mean(self, in_, mean):
        """
        Set the mean to subtract for centering the data.

        Parameters
        ----------
        in_ : which input to assign this mean.
        mean : mean ndarray (input dimensional or broadcastable)
        """
        self.__check_input(in_)
        ms = mean.shape
        if mean.ndim == 1:
            # broadcast channels
            if ms[0] != self.inputs[in_][1]:
                raise ValueError('Mean channels incompatible with input.')
            mean = mean[:, np.newaxis, np.newaxis]
        else:
            # elementwise mean
            if len(ms) == 2:
                ms = (1,) + ms
            if len(ms) != 3:
                raise ValueError('Mean shape invalid')
            if ms != self.inputs[in_][1:]:
                in_shape = self.inputs[in_][1:]
                m_min, m_max = mean.min(), mean.max()
                normal_mean = (mean - m_min) / (m_max - m_min)
                mean = resize_image(normal_mean.transpose((1,2,0)),in_shape[1:]).transpose((2,0,1)) * (m_max - m_min) + m_min
                #raise ValueError('Mean shape incompatible with input shape.')
        self.mean[in_] = mean

def set_mean(self, in_, mean):

"""

Set the mean to subtract for centering the data.

Parameters

----------

in_ : which input to assign this mean.

mean : mean ndarray (input dimensional or broadcastable)

"""

self.__check_input(in_)

ms = mean.shape

if mean.ndim == 1:

# broadcast channels

if ms[0] != self.inputs[in_][1]:

raise ValueError('Mean channels incompatible with input.')

mean = mean[:, np.newaxis, np.newaxis]

else:

# elementwise mean

if len(ms) == 2:

ms = (1,) + ms

if len(ms) != 3:

raise ValueError('Mean shape invalid')

if ms != self.inputs[in_][1:]:

in_shape = self.inputs[in_][1:]

m_min, m_max = mean.min(), mean.max()

normal_mean = (mean - m_min) / (m_max - m_min)

mean = resize_image(normal_mean.transpose((1,2,0)),in_shape[1:]).transpose((2,0,1)) * (m_max - m_min) + m_min

#raise ValueError('Mean shape incompatible with input shape.')

self.mean[in_] = mean

调整完成后，需要重新编译Caffe:

$ make clean
$ make
$ make pycaffe
$ make distribute

$ make clean

$ make

$ make pycaffe

$ make distribute

参考链接

macOS Sierra (10.12.4)下Caffe通过Python接口加载binaryproto格式的均值文件

macOS Sierra (10.12.4)下Caffe通过Python接口加载均值文件的时候，都是加载的.npy格式的文件，这个格式是Python存储的格式，跟我们经常下载到的.binaryproto格式的均值文件是不同的，这样就导致了加载问题。
.binaryproto是Google的Protocol Buffer序列化后的数据，而.npy格式是Python的numpy模块序列化后的数据。

之所以会出现两种不同的存储格式，目前猜测是由于目前Python 3不能很好的支持Protocol Buffer导致的。

Python下是不能直接加载.binaryproto格式的数据的，必须进行一次转换才行，示例代码如下：

#coding=utf-8
#加载必要的库
import numpy as np
 
import sys,os

#设置当前目录
caffe_root = '/Users/Source/caffe/distribute/'
sys.path.append(caffe_root + 'python')

import caffe
os.chdir(caffe_root)

memnet_root = '/Users/Source/caffe/'
model_def =memnet_root + 'models/memnet/deploy.prototxt'
pretrained_model=memnet_root + 'models/memnet/memnet.caffemodel'
means_file=memnet_root + 'models/memnet/mean.binaryproto'

caffe.set_mode_cpu() 
blobProto=caffe.io.caffe_pb2.BlobProto()
binProtoFile=open(means_file,'rb')
blobProto.ParseFromString(binProtoFile.read())
means = caffe.io.blobproto_to_array(blobProto)[0]
binProtoFile.close()

# Make detector.
detector = caffe.Detector(model_def, pretrained_model, mean=means)

#coding=utf-8

#加载必要的库

import numpy as np

import sys,os

#设置当前目录

caffe_root = '/Users/Source/caffe/distribute/'

sys.path.append(caffe_root + 'python')

import caffe

os.chdir(caffe_root)

memnet_root = '/Users/Source/caffe/'

model_def =memnet_root + 'models/memnet/deploy.prototxt'

pretrained_model=memnet_root + 'models/memnet/memnet.caffemodel'

means_file=memnet_root + 'models/memnet/mean.binaryproto'

caffe.set_mode_cpu()

blobProto=caffe.io.caffe_pb2.BlobProto()

binProtoFile=open(means_file,'rb')

blobProto.ParseFromString(binProtoFile.read())

means = caffe.io.blobproto_to_array(blobProto)[0]

binProtoFile.close()

# Make detector.

detector = caffe.Detector(model_def, pretrained_model, mean=means)

参考链接

macOS Sierra (10.12.4)编译pycaffe成功后，执行时候崩溃，错误“Segmentation fault: 11”

参照 macOS Sierra (10.12.3)编译Caffe 编译成功 Caffe 后,开始尝试使用 Caffe 的 Python 接口，执行如下命令：

$ make pycaffe

1	$ make pycaffe

编译一切成功，但是当执行

import caffe

1	import caffe

的时候，程序崩溃，提示如下内容：

Segmentation fault: 11

1	Segmentation fault: 11

继续阅读

macOS Sierra (10.12.4)下Python通过PyAV调用FFMPEG操作视频

macOS Sierra (10.12.4)下使用Python操作视频，FFMPEG是目前来说最好的一个选择，但是没有为Python专门提供适配接口，网上搜索了比较长时间，才找到PyAV来操作FFMPEG。

PyAV的文档地址在：https://mikeboers.github.io/PyAV/

代码地址在：https://github.com/mikeboers/PyAV

首先需要通过HomeBrew安装FFMPEG：

$ brew install ffmpeg

1	$ brew install ffmpeg

接下来安装PyAV，安装方式两种：

一种是直接通过PIP来安装：

$ pip install av

1	$ pip install av

另外一种是通过下载代码来手工安装

$ git clone https://github.com/mikeboers/PyAV.git
$ cd PyAV
$ python setup.py install

$ git clone https://github.com/mikeboers/PyAV.git

$ cd PyAV

$ python setup.py install

安装好后的例子如下：

import av
from av.frame import Frame
from av.packet import Packet
from av.stream import Stream
from av.utils import AVError
from av.video import VideoFrame

container = av.open('san.mp4')

for frame in container.decode(video=0):
    frame.to_image().save('./pyav/frame-%04d.jpg' % frame.index)

import av

from av.frame import Frame

from av.packet import Packet

from av.stream import Stream

from av.utils import AVError

from av.video import VideoFrame

container = av.open('san.mp4')

for frame in container.decode(video=0):

frame.to_image().save('./pyav/frame-%04d.jpg' % frame.index)

一	二	三	四	五	六	日
	1	2	3	4	5	6
7	8	9	10	11	12	13
14	15	16	17	18	19	20
21	22	23	24	25	26	27
28	29	30