# Save PointMap to pm_path. def SavePlyFile(self,pm_path, pm, image): width = pm.GetSize().width height = pm.GetSize().height pm_num = width * height pm_ptr = np.array(pm, copy=False).reshape((-1, 3)) image_ptr = np.array(image, copy=False) if image.GetType() == RVC.ImageTypeEnum.Mono8: tmp = image_ptr.copy().flatten() image_ptr = np.zeros((height * width, 3)) image_ptr[:, 0] = tmp image_ptr[:, 1] = tmp image_ptr[:, 2] = tmp else: image_ptr = image_ptr.reshape(-1, 3) if image.GetType() == RVC.ImageTypeEnum.BGR8: image_ptr[:, [0, 2]] = image_ptr[:, [2, 0]] data = np.concatenate((pm_ptr, image_ptr), axis=1) np.savetxt(pm_path, data, fmt = '%f %f %f %d %d %d') with open(pm_path, 'r+') as f: old = f.read() f.seek(0) f.write("{}".format("ply")) f.write("\n{}".format("format ascii 1.0")) f.write("\n{}".format("comment Created by Rvbust, Inc")) f.write("\nelement vertex {}".format(pm_num)) f.write("\n{}".format("property float x")) f.write("\n{}".format("property float y")) f.write("\n{}".format("property float z")) f.write("\n{}".format("property uchar red")) f.write("\n{}".format("property uchar green")) f.write("\n{}".format("property uchar blue")) f.write("\n{}\n".format("end_header")) f.write(old) f.close() return
时间: 2024-02-14 18:24:04 浏览: 21
这段代码是用来将一个 PointMap 和一张图片保存为一个 .ply 文件的函数。具体来说,它接受三个参数:pm_path 表示保存的路径,pm 表示 PointMap 对象,image 表示图片对象。
首先,它通过 pm.GetSize() 获取 PointMap 的宽高信息,计算出 PointMap 中点的数量。然后,它将 PointMap 转换为一个二维 numpy 数组 pm_ptr,将图片转换为一个三维 numpy 数组 image_ptr。
接下来,它根据图片的类型进行判断。如果是 Mono8 类型,表示图片是单通道灰度图像,那么它将其转换为一个三通道数组,其中三个通道的值都相等。如果是其他类型,那么就将其转换为一个二维的三通道数组,并将 BGR 通道顺序转换为 RGB。
最后,它将 pm_ptr 和 image_ptr 按列连接成一个大的 numpy 数组 data,并将其写入到 pm_path 指定的文件中。在写入之前,它先写入了一个 .ply 文件的头部信息,包括文件格式、版本、注释、点的数量和属性等。最后,它将原文件中的内容追加到新写入的内容之后,完成文件的保存。
相关问题
import os from PyQt5.QtCore import Qt from PyQt5.QtGui import QPixmap, QIcon from PyQt5.QtWidgets import QApplication, QWidget, QLabel, QVBoxLayout, QHBoxLayout, QTreeView, QFileSystemModel class ImageViewer(QWidget): def init(self, folder_path): super().init() self.folder_path = folder_path self.image_dict = {} self.current_image = None self.setWindowTitle("Image Viewer") self.setFixedSize(1000, 600) self.image_label = QLabel(self) self.image_label.setAlignment(Qt.AlignCenter) self.tree_view = QTreeView() self.tree_view.setMinimumWidth(250) self.tree_view.setMaximumWidth(250) self.model = QFileSystemModel() self.model.setRootPath(folder_path) self.tree_view.setModel(self.model) self.tree_view.setRootIndex(self.model.index(folder_path)) self.tree_view.setHeaderHidden(True) self.tree_view.setColumnHidden(1, True) self.tree_view.setColumnHidden(2, True) self.tree_view.setColumnHidden(3, True) self.tree_view.doubleClicked.connect(self.tree_item_double_clicked) self.main_layout = QHBoxLayout(self) self.main_layout.addWidget(self.tree_view) self.main_layout.addWidget(self.image_label) self.load_images() self.update_image() def load_images(self): for file_name in os.listdir(self.folder_path): if file_name.lower().endswith((".jpg", ".jpeg", ".png", ".gif", ".bmp")): file_path = os.path.join(self.folder_path, file_name) self.image_dict[file_name] = file_path current_image = list(self.image_dict.keys())[0] def update_image(self): if self.current_image is not None: pixmap = QPixmap(self.image_dict[self.current_image]) self.image_label.setPixmap(pixmap.scaled(self.width() - self.tree_view.width(), self.height(), Qt.KeepAspectRatio, Qt.SmoothTransformation)) def tree_item_double_clicked(self, index): file_name = self.model.fileName(index) if file_name in self.image_dict: self.current_image = file_name self.update_image() def keyPressEvent(self, event): if event.key() == Qt.Key_A: self.previous_image() elif event.key() == Qt.Key_D: self.next_image() elif event.key() in [Qt.Key_1, Qt.Key_2, Qt.Key_3, Qt.Key_4, Qt.Key_5]: self.save_text_file(event.key() - Qt.Key_0) def previous_image(self): if self.current_image is not None: file_names = list(self.image_dict.keys()) current_index = file_names.index(self.current_image) if current_index > 0: self.current_image = file_names[current_index - 1] else: self.current_image = file_names[-1] self.update_image() def next_image(self): if self.current_image is not None: file_names = list(self.image_dict.keys()) current_index = file_names.index(self.current_image) if current_index < len(file_names) - 1: self.current_image = file_names[current_index + 1] else: self.current_image = file_names[0] self.update_image() def save_text_file(self, number): if self.current_image is not None: file_name = self.current_image txt_file_path = os.path.join(self.folder_path, os.path.splitext(file_name)[0] + ".txt") with open(txt_file_path, "w") as file: file.write(str(number)) if name == "main": import sys app = QApplication(sys.argv) viewer = ImageViewer("D:/图片/wallpaper") viewer.show() sys.exit(app.exec_())这份代码实现不了使用键盘的A键向上翻页以及D键向下翻页,也实现不了键盘数字键生成相应txt文档,帮我分析一下错在哪里
这份代码在实现键盘事件的时候,使用了错误的函数名称。应该使用`keyPressEvent`而不是`key_press_event`函数。另外,在`save_text_file`函数中,应该先判断当前图片是否为`None`,再执行后续的代码。以下是修改后的代码:
```
import os
from PyQt5.QtCore import Qt
from PyQt5.QtGui import QPixmap, QIcon
from PyQt5.QtWidgets import QApplication, QWidget, QLabel, QVBoxLayout, QHBoxLayout, QTreeView, QFileSystemModel
class ImageViewer(QWidget):
def __init__(self, folder_path):
super().__init__()
self.folder_path = folder_path
self.image_dict = {}
self.current_image = None
self.setWindowTitle("Image Viewer")
self.setFixedSize(1000, 600)
self.image_label = QLabel(self)
self.image_label.setAlignment(Qt.AlignCenter)
self.tree_view = QTreeView()
self.tree_view.setMinimumWidth(250)
self.tree_view.setMaximumWidth(250)
self.model = QFileSystemModel()
self.model.setRootPath(folder_path)
self.tree_view.setModel(self.model)
self.tree_view.setRootIndex(self.model.index(folder_path))
self.tree_view.setHeaderHidden(True)
self.tree_view.setColumnHidden(1, True)
self.tree_view.setColumnHidden(2, True)
self.tree_view.setColumnHidden(3, True)
self.tree_view.doubleClicked.connect(self.tree_item_double_clicked)
self.main_layout = QHBoxLayout(self)
self.main_layout.addWidget(self.tree_view)
self.main_layout.addWidget(self.image_label)
self.load_images()
self.update_image()
def load_images(self):
for file_name in os.listdir(self.folder_path):
if file_name.lower().endswith((".jpg", ".jpeg", ".png", ".gif", ".bmp")):
file_path = os.path.join(self.folder_path, file_name)
self.image_dict[file_name] = file_path
self.current_image = list(self.image_dict.keys())[0]
def update_image(self):
if self.current_image is not None:
pixmap = QPixmap(self.image_dict[self.current_image])
self.image_label.setPixmap(pixmap.scaled(self.width() - self.tree_view.width(), self.height(), Qt.KeepAspectRatio, Qt.SmoothTransformation))
def tree_item_double_clicked(self, index):
file_name = self.model.fileName(index)
if file_name in self.image_dict:
self.current_image = file_name
self.update_image()
def keyPressEvent(self, event):
if event.key() == Qt.Key_A:
self.previous_image()
elif event.key() == Qt.Key_D:
self.next_image()
elif event.key() in [Qt.Key_1, Qt.Key_2, Qt.Key_3, Qt.Key_4, Qt.Key_5]:
self.save_text_file(event.key() - Qt.Key_0)
def previous_image(self):
if self.current_image is not None:
file_names = list(self.image_dict.keys())
current_index = file_names.index(self.current_image)
if current_index > 0:
self.current_image = file_names[current_index - 1]
else:
self.current_image = file_names[-1]
self.update_image()
def next_image(self):
if self.current_image is not None:
file_names = list(self.image_dict.keys())
current_index = file_names.index(self.current_image)
if current_index < len(file_names) - 1:
self.current_image = file_names[current_index + 1]
else:
self.current_image = file_names[0]
self.update_image()
def save_text_file(self, number):
if self.current_image is not None:
file_name = self.current_image
txt_file_path = os.path.join(self.folder_path, os.path.splitext(file_name)[0] + ".txt")
with open(txt_file_path, "w") as file:
file.write(str(number))
if __name__ == "__main__":
import sys
app = QApplication(sys.argv)
viewer = ImageViewer("D:/图片/wallpaper")
viewer.show()
sys.exit(app.exec_())
```
这段代码中加一个test loss功能 class LSTM(nn.Module): def __init__(self, input_size, hidden_size, num_layers, output_size, batch_size, device): super().__init__() self.device = device self.input_size = input_size self.hidden_size = hidden_size self.num_layers = num_layers self.output_size = output_size self.num_directions = 1 # 单向LSTM self.batch_size = batch_size self.lstm = nn.LSTM(self.input_size, self.hidden_size, self.num_layers, batch_first=True) self.linear = nn.Linear(65536, self.output_size) def forward(self, input_seq): h_0 = torch.randn(self.num_directions * self.num_layers, self.batch_size, self.hidden_size).to(self.device) c_0 = torch.randn(self.num_directions * self.num_layers, self.batch_size, self.hidden_size).to(self.device) output, _ = self.lstm(input_seq, (h_0, c_0)) pred = self.linear(output.contiguous().view(self.batch_size, -1)) return pred if __name__ == '__main__': # 加载已保存的模型参数 saved_model_path = '/content/drive/MyDrive/危急值/model/dangerous.pth' device = 'cuda:0' lstm_model = LSTM(input_size=1, hidden_size=64, num_layers=1, output_size=3, batch_size=256, device='cuda:0').to(device) state_dict = torch.load(saved_model_path) lstm_model.load_state_dict(state_dict) dataset = ECGDataset(X_train_df.to_numpy()) dataloader = DataLoader(dataset, batch_size=256, shuffle=True, num_workers=0, drop_last=True) loss_fn = nn.CrossEntropyLoss() optimizer = optim.SGD(lstm_model.parameters(), lr=1e-4) for epoch in range(200000): print(f'epoch:{epoch}') lstm_model.train() epoch_bar = tqdm(dataloader) for x, y in epoch_bar: optimizer.zero_grad() x_out = lstm_model(x.to(device).type(torch.cuda.FloatTensor)) loss = loss_fn(x_out, y.long().to(device)) loss.backward() epoch_bar.set_description(f'loss:{loss.item():.4f}') optimizer.step() if epoch % 100 == 0 or epoch == epoch - 1: torch.save(lstm_model.state_dict(), "/content/drive/MyDrive/危急值/model/dangerous.pth") print("权重成功保存一次")
以下是加入 Test Loss 功能后的代码:
```python
import torch
import torch.nn as nn
import torch.optim as optim
from torch.utils.data import DataLoader
class LSTM(nn.Module):
def __init__(self, input_size, hidden_size, num_layers, output_size, batch_size, device):
super().__init__()
self.device = device
self.input_size = input_size
self.hidden_size = hidden_size
self.num_layers = num_layers
self.output_size = output_size
self.num_directions = 1 # 单向LSTM
self.batch_size = batch_size
self.lstm = nn.LSTM(self.input_size, self.hidden_size, self.num_layers, batch_first=True)
self.linear = nn.Linear(65536, self.output_size)
def forward(self, input_seq):
h_0 = torch.randn(self.num_directions * self.num_layers, self.batch_size, self.hidden_size).to(self.device)
c_0 = torch.randn(self.num_directions * self.num_layers, self.batch_size, self.hidden_size).to(self.device)
output, _ = self.lstm(input_seq, (h_0, c_0))
pred = self.linear(output.contiguous().view(self.batch_size, -1))
return pred
if __name__ == '__main__':
# 加载已保存的模型参数
saved_model_path = '/content/drive/MyDrive/危急值/model/dangerous.pth'
device = 'cuda:0'
lstm_model = LSTM(input_size=1, hidden_size=64, num_layers=1, output_size=3, batch_size=256, device=device).to(device)
state_dict = torch.load(saved_model_path)
lstm_model.load_state_dict(state_dict)
# 加载测试数据集
test_dataset = ECGDataset(X_test_df.to_numpy())
test_loader = DataLoader(test_dataset, batch_size=256, shuffle=False, num_workers=0, drop_last=False)
# 定义损失函数和优化器
loss_fn = nn.CrossEntropyLoss()
optimizer = optim.SGD(lstm_model.parameters(), lr=1e-4)
# 开始训练
for epoch in range(200000):
print(f'epoch:{epoch}')
lstm_model.train()
epoch_bar = tqdm(dataloader)
for x, y in epoch_bar:
optimizer.zero_grad()
x_out = lstm_model(x.to(device).type(torch.cuda.FloatTensor))
loss = loss_fn(x_out, y.long().to(device))
loss.backward()
epoch_bar.set_description(f'loss:{loss.item():.4f}')
optimizer.step()
# 计算测试集上的平均损失
test_loss = 0.0
lstm_model.eval()
with torch.no_grad():
for x, y in test_loader:
x_out = lstm_model(x.to(device).type(torch.cuda.FloatTensor))
loss = loss_fn(x_out, y.long().to(device))
test_loss += loss.item() * x.size(0)
test_loss /= len(test_dataset)
print(f'Test Loss: {test_loss:.4f}')
if epoch % 100 == 0 or epoch == epoch - 1:
torch.save(lstm_model.state_dict(), "/content/drive/MyDrive/危急值/model/dangerous.pth")
print("权重成功保存一次")
```
在这个代码中,我们首先加载了测试数据集 `test_dataset` 和测试数据加载器 `test_loader`。在每个 epoch 完成后,我们计算测试集上的平均损失值。我们使用 `lstm_model.eval()` 将模型设为评估模式,并利用 `torch.no_grad()` 避免计算梯度,以加速计算。最后,我们输出测试集上的平均损失值。