主页 / user-235210

Max's questions

Martin Hope
Max
Asked: 2020-09-10 22:21:59 +0000 UTC
  • 10

您好，任务是在python中使用Keras，theano，使用完全卷积神经网络搜索对象。当前使用以下网络配置：

def create_encoding_layers():
    kernel = 3
    filter_size = 64
    pad = 1
    pool_size = 2
    return [
        ZeroPadding2D(padding=(pad,pad)),
        Convolution2D(filter_size, kernel, kernel, border_mode='valid'),
        BatchNormalization(),
        Activation('relu'),
        MaxPooling2D(pool_size=(pool_size, pool_size)),

        ZeroPadding2D(padding=(pad,pad)),
        Convolution2D(32, kernel, kernel, border_mode='valid'),
        BatchNormalization(),
        Activation('relu'),
        MaxPooling2D(pool_size=(pool_size, pool_size)),

        ZeroPadding2D(padding=(pad,pad)),
        Convolution2D(32, kernel, kernel, border_mode='valid'),
        BatchNormalization(),
        Activation('relu'),
        MaxPooling2D(pool_size=(pool_size, pool_size)),

        ZeroPadding2D(padding=(pad,pad)),
        Convolution2D(64, kernel, kernel, border_mode='valid'),
        BatchNormalization(),
        Activation('relu'),
    ]
    def create_decoding_layers():
    kernel = 3
    filter_size = 64
    pad = 1
    pool_size = 2
    return[
        ZeroPadding2D(padding=(pad,pad)),
        Convolution2D(64, kernel, kernel, border_mode='valid'),
        BatchNormalization(),

        UpSampling2D(size=(pool_size,pool_size)),
        ZeroPadding2D(padding=(pad,pad)),
        Convolution2D(32, kernel, kernel, border_mode='valid'),
        BatchNormalization(),

        UpSampling2D(size=(pool_size,pool_size)),
        ZeroPadding2D(padding=(pad,pad)),
        Convolution2D(32, kernel, kernel, border_mode='valid'),
        BatchNormalization(),

        UpSampling2D(size=(pool_size,pool_size)),
        ZeroPadding2D(padding=(pad,pad)),
        Convolution2D(filter_size, kernel, kernel,     border_mode='valid'),
        BatchNormalization(),
        ZeroPadding2D(padding=(0,pad+1)),
    ]

    segnet_basic = models.Sequential()
    segnet_basic.add(Layer(input_shape=(120, 420, 1)))
    segnet_basic.encoding_layers = create_encoding_layers()
    for l in segnet_basic.encoding_layers:
        segnet_basic.add(l)
    segnet_basic.decoding_layers = create_decoding_layers()
    for l in segnet_basic.decoding_layers:
        segnet_basic.add(l)
    segnet_basic.add(Convolution2D(1, 1, 1, border_mode='valid',))
    segnet_basic.summary()
    segnet_basic.add(Reshape((120,420,1)))

输入是一张照片和一个蒙版（对象所在的区域用 1 填充，其余区域用 0 填充）。据我了解，在输出中我们应该得到相同的掩码。尝试了不同的损失函数。输出是废话。实际上问题是：为了在输出中获得类似的掩码，需要纠正什么？使用哪个损失函数和优化器？图片显示了一个已经训练好的网络的工作示例。image - 输入网络的图像，label - 预期结果，我乘以 255 作为输出，result - 网络产生的结果。我还乘以 255 作为输出 在此处输入图像描述

python
Martin Hope
Max
Asked: 2020-06-26 20:41:17 +0000 UTC
  • 4

执行代码后，我得到一张黑白图像，它由两个维度（宽度高度）和值本身组成，从0到255- 从黑色到白色。为了将来使用，我需要将其转换为rgb8. 例如，该功能cv2.floodFill不接受我收到的图像。这是代码本身：

import sys
import cv2
import numpy as np
import time
from matplotlib import pyplot as plt

img = cv2.imread('2.jpg')
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
ret, thresh = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY_INV+cv2.THRESH_OTSU)
# noise removal
kernel = np.ones((3, 3), np.uint8)
opening = cv2.morphologyEx(thresh, cv2.MORPH_OPEN, kernel, iterations = 2)
# sure background area
sure_bg = cv2.dilate(opening, kernel, iterations=3)
# Finding sure foreground area
dist_transform = cv2.distanceTransform(opening, cv2.DIST_L2, 5)
ret, sure_fg = cv2.threshold(dist_transform, 0.55 * dist_transform.max(), 255, 0)
# Finding unknown region
sure_fg = np.uint8(sure_fg)
unknown = cv2.subtract(sure_bg, sure_fg)
# Marker labelling
ret, markers = cv2.connectedComponents(sure_fg)
# Add one to all labels so that sure background is not 0, but 1

markers = markers+10
# Now, mark the region of unknown with zero
markers[unknown == 255] = 0
markers = cv2.watershed(img, markers)
img[markers == -1] = [255, 0, 0]
cv2.imwrite('6.jpg', markers)

是否可以转换markers为普通图像？

python
Martin Hope
Max
Asked: 2020-02-07 04:33:47 +0000 UTC
  • 0

无法连接到树莓派上的无线网络。网络已定位，但当您尝试连接到网络时，已建立连接，但连接立即断开。我使用 wic 实用程序。覆盆子系统。ifconfig 结果

eth0      Link encap:Ethernet  HWaddr b8:27:eb:ba:a4:1a  
          UP BROADCAST MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)

lo        Link encap:Local Loopback  
          inet addr:127.0.0.1  Mask:255.0.0.0
          inet6 addr: ::1/128 Scope:Host
          UP LOOPBACK RUNNING  MTU:65536  Metric:1
          RX packets:328 errors:0 dropped:0 overruns:0 frame:0
          TX packets:328 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1 
          RX bytes:26668 (26.0 KiB)  TX bytes:26668 (26.0 KiB)

usb0      Link encap:Ethernet  HWaddr 02:68:04:60:69:78  
          inet addr:192.168.42.39  Bcast:192.168.42.255  Mask:255.255.255.0
          inet6 addr: fe80::b3f6:1e61:38e2:eaa7/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:3632 errors:0 dropped:0 overruns:0 frame:0
          TX packets:3147 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:3492653 (3.3 MiB)  TX bytes:498164 (486.4 KiB)

wlan0     Link encap:Ethernet  HWaddr b8:27:eb:ef:f1:4f  
          UP BROADCAST MULTICAST  MTU:1500  Metric:1
          RX packets:105 errors:0 dropped:105 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:7003 (6.8 KiB)  TX bytes:0 (0.0 B)

Feb 7 02:17:31 raspberrypi kernel: [ 2171.438884] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready Feb 7 02:17:31 raspberrypi kernel: [ 2171.439000] brcmfmac: power management disabled Feb 7 02:17:42 raspberrypi kernel: [ 2181.710227] smsc95xx 1-1.1:1.0 eth0: hardware isn't capable of remote wakeup Feb 7 02:17:42 raspberrypi kernel: [ 2181.781102] smsc95xx 1-1.1:1.0 eth0: hardware isn't capable of remote wakeup Feb 7 02:17:42 raspberrypi kernel: [ 2181.781816] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready Feb 7 02:17:43 raspberrypi kernel: [ 2182.531561] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready Feb 7 02:17:43 raspberrypi kernel: [ 2182.531627] brcmfmac: power management disabled Feb 7 02:17:46 raspberrypi kernel: [ 2186.218582] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready Feb 7 02:17:46 raspberrypi kernel: [ 2186.277077] brcmfmac: power management disabled Feb 7 02:17:46 raspberrypi kernel: [ 2186.484581] smsc95xx 1-1.1:1.0 eth0: hardware isn't capable of remote wakeup Feb 7 02:17:47 raspberrypi kernel: [ 2186.560969] smsc95xx 1-1.1:1.0 eth0: hardware isn't capable of remote wakeup Feb 7 02:17:47 raspberrypi kernel: [ 2186.561417] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready

linux
Martin Hope
Max
Asked: 2020-01-27 05:48:07 +0000 UTC
  • 2

需要通过控制台来组织键盘的处理。你需要分配一个动作来按下按钮并按下它。例如，pygame 解决了一个类似的任务，它看起来像这样：

for events in pygame.event.get():
    if events.type == pygame.KEYDOWN:
        #проверяем конкретную клавишу...
    if events.type == pygame.KEYUP:
        #проверяем конкретную клавишу...

但在 pygame 中，捕获发生在创建的窗口中，而不是在控制台中。

python