【Python】圆柱体内部3D点云仿真及ply文件生成，圆形3D点云检测及拟合算法
更新以gitee为准：
管道内管壁ply点云生成及管道检测
文章目录
ply点云显示
通过plyfile库可以完成点云的读取或写入
同时 引入color map 使其与不同的深度值进行对应
ply读取函数如下：
python
color_map = [
(128, 0, 0), (132, 0, 0), (136, 0, 0), (140, 0, 0),
(144, 0, 0), (148, 0, 0), (152, 0, 0), (156, 0, 0),
(160, 0, 0), (164, 0, 0), (168, 0, 0), (172, 0, 0),
(176, 0, 0), (180, 0, 0), (184, 0, 0), (188, 0, 0),
(192, 0, 0), (196, 0, 0), (200, 0, 0), (204, 0, 0),
(208, 0, 0), (212, 0, 0), (216, 0, 0), (220, 0, 0),
(224, 0, 0), (228, 0, 0), (232, 0, 0), (236, 0, 0),
(240, 0, 0), (244, 0, 0), (248, 0, 0), (252, 0, 0),
(255, 0, 0), (255, 4, 0), (255, 8, 0), (255, 12, 0),
(255, 16, 0), (255, 20, 0), (255, 24, 0), (255, 28, 0),
(255, 32, 0), (255, 36, 0), (255, 40, 0), (255, 44, 0),
(255, 48, 0), (255, 52, 0), (255, 56, 0), (255, 60, 0),
(255, 64, 0), (255, 68, 0), (255, 72, 0), (255, 76, 0),
(255, 80, 0), (255, 84, 0), (255, 88, 0), (255, 92, 0),
(255, 96, 0), (255, 100, 0), (255, 104, 0), (255, 108, 0),
(255, 112, 0), (255, 116, 0), (255, 120, 0), (255, 124, 0),
(255, 128, 0), (255, 132, 0), (255, 136, 0), (255, 140, 0),
(255, 144, 0), (255, 148, 0), (255, 152, 0), (255, 156, 0),
(255, 160, 0), (255, 164, 0), (255, 168, 0), (255, 172, 0),
(255, 176, 0), (255, 180, 0), (255, 184, 0), (255, 188, 0),
(255, 192, 0), (255, 196, 0), (255, 200, 0), (255, 204, 0),
(255, 208, 0), (255, 212, 0), (255, 216, 0), (255, 220, 0),
(255, 224, 0), (255, 228, 0), (255, 232, 0), (255, 236, 0),
(255, 240, 0), (255, 244, 0), (255, 248, 0), (255, 252, 0),
(254, 255, 1), (250, 255, 6), (246, 255, 10), (242, 255, 14),
(238, 255, 18), (234, 255, 22), (230, 255, 26), (226, 255, 30),
(222, 255, 34), (218, 255, 38), (214, 255, 42), (210, 255, 46),
(206, 255, 50), (202, 255, 54), (198, 255, 58), (194, 255, 62),
(190, 255, 66), (186, 255, 70), (182, 255, 74), (178, 255, 78),
(174, 255, 82), (170, 255, 86), (166, 255, 90), (162, 255, 94),
(158, 255, 98), (154, 255, 102), (150, 255, 106), (146, 255, 110),
(142, 255, 114), (138, 255, 118), (134, 255, 122), (130, 255, 126),
(126, 255, 130), (122, 255, 134), (118, 255, 138), (114, 255, 142),
(110, 255, 146), (106, 255, 150), (102, 255, 154), (98, 255, 158),
(94, 255, 162), (90, 255, 166), (86, 255, 170), (82, 255, 174),
(78, 255, 178), (74, 255, 182), (70, 255, 186), (66, 255, 190),
(62, 255, 194), (58, 255, 198), (54, 255, 202), (50, 255, 206),
(46, 255, 210), (42, 255, 214), (38, 255, 218), (34, 255, 222),
(30, 255, 226), (26, 255, 230), (22, 255, 234), (18, 255, 238),
(14, 255, 242), (10, 255, 246), (6, 255, 250), (2, 255, 254),
(0, 252, 255), (0, 248, 255), (0, 244, 255), (0, 240, 255),
(0, 236, 255), (0, 232, 255), (0, 228, 255), (0, 224, 255),
(0, 220, 255), (0, 216, 255), (0, 212, 255), (0, 208, 255),
(0, 204, 255), (0, 200, 255), (0, 196, 255), (0, 192, 255),
(0, 188, 255), (0, 184, 255), (0, 180, 255), (0, 176, 255),
(0, 172, 255), (0, 168, 255), (0, 164, 255), (0, 160, 255),
(0, 156, 255), (0, 152, 255), (0, 148, 255), (0, 144, 255),
(0, 140, 255), (0, 136, 255), (0, 132, 255), (0, 128, 255),
(0, 124, 255), (0, 120, 255), (0, 116, 255), (0, 112, 255),
(0, 108, 255), (0, 104, 255), (0, 100, 255), (0, 96, 255),
(0, 92, 255), (0, 88, 255), (0, 84, 255), (0, 80, 255),
(0, 76, 255), (0, 72, 255), (0, 68, 255), (0, 64, 255),
(0, 60, 255), (0, 56, 255), (0, 52, 255), (0, 48, 255),
(0, 44, 255), (0, 40, 255), (0, 36, 255), (0, 32, 255),
(0, 28, 255), (0, 24, 255), (0, 20, 255), (0, 16, 255),
(0, 12, 255), (0, 8, 255), (0, 4, 255), (0, 0, 255),
(0, 0, 252), (0, 0, 248), (0, 0, 244), (0, 0, 240),
(0, 0, 236), (0, 0, 232), (0, 0, 228), (0, 0, 224),
(0, 0, 220), (0, 0, 216), (0, 0, 212), (0, 0, 208),
(0, 0, 204), (0, 0, 200), (0, 0, 196), (0, 0, 192),
(0, 0, 188), (0, 0, 184), (0, 0, 180), (0, 0, 176),
(0, 0, 172), (0, 0, 168), (0, 0, 164), (0, 0, 160),
(0, 0, 156), (0, 0, 152), (0, 0, 148), (0, 0, 144),
(0, 0, 140), (0, 0, 136), (0, 0, 132), (0, 0, 128)
]
def visualize_ply(file_path):
# 读取PLY文件
ply_data = PlyData.read(file_path)
vertices = ply_data['vertex']
# 提取坐标数据
x = vertices['x']
y = vertices['y']
z = vertices['z']
# 计算深度范围
min_depth = np.min(z)
max_depth = np.max(z)
depth_range = max_depth - min_depth
# 生成颜色映射
colors = []
for depth in z:
# 将深度映射到颜色索引 (0-255)
color_idx = int(((depth - min_depth) / depth_range) * (len(color_map)-1))
colors.append(color_map[color_idx])
# 转换为0-1范围的RGB值
colors = np.array(colors) / 255.0
# 创建3D可视化
fig = plt.figure(figsize=(10, 8))
ax = fig.add_subplot(111, projection='3d')
# 绘制点云
ax.scatter(x, y, z, c=colors, s=1, marker='o')
# 设置坐标轴标签
ax.set_xlabel('X')
ax.set_ylabel('Y')
ax.set_zlabel('Z')
# 启用交互式旋转
plt.show(block=True) # 添加block=True参数保持窗口显示点云生成
建立一个tkinter界面 用于生成
其中 fov角度的引入表示 某一传感器防止在圆柱体内部中心位置 延伸z轴的正向延长方向上做检测 传感器测量范围有限 那么最近的一个点距离S与fov的关系为：
sin(fov/2)=R/S
python
min_range = radius / math.sin(math.radians(fov_deg / 2))即最大fov角度时 测到的管壁的第一个点 其必须小于测量范围 否则就完全测不到
对应的z轴最小范围为：
python
min_distance = radius / math.tan(math.radians(fov_deg / 2))当测得最远的一个点时 其斜边为最大测量范围 那么此时对应的角度a与半径关系为：
sin(a)=R/max_range
求得的a要小于fov/2才行
根据几何关系如果第一个点能被测得到 那么距离最大的一个点肯定也能被侧得到
但也可加上条件判断
python
a_rad = math.asin(radius / max_range)
a_deg = math.degrees(a_rad)
if a_deg > fov_deg / 2:
return None此时 距离最大值是斜边 半径为对边 z轴最大距离s为临边
那么z轴最大距离即为：
python
max_distance = math.sqrt(max_range**2 - radius**2)最后再加上点的数量、噪音、以及近距离分布多 远距离分布少的系数得到
生成函数为：
python
def generate_cylinder(radius, fov_deg, max_range, points=5000,noise=5,coefficient=4):
if radius <= 0 or fov_deg <= 0 or max_range <= 0:
return None
if fov_deg > 180:
return None
if radius >= max_range:
return None
if points <= 0:
return None
min_range = radius / math.sin(math.radians(fov_deg / 2))
if min_range > max_range:
return None
max_distance = math.sqrt(max_range**2 - radius**2)
min_distance = radius / math.tan(math.radians(fov_deg / 2))
if max_distance <= min_distance or max_distance <= 0 or min_distance <= 0:
return None
a_rad = math.asin(radius / max_range)
a_deg = math.degrees(a_rad)
if a_deg > fov_deg / 2:
return None
# 生成圆柱面点云
theta = np.random.uniform(0, 2*math.pi, points)
u = np.random.rand(points)
z = max_distance - (max_distance - min_distance) * np.power(u,1/coefficient)
# z = np.random.uniform(min_distance, max_distance, points)
# 转换为笛卡尔坐标（管壁表面）
radius_noise = np.random.normal(0, noise, points)
radius += radius_noise
x = radius * np.cos(theta)
y = radius * np.sin(theta)
x_noise = np.random.normal(0, noise, points)
y_noise = np.random.normal(0, noise, points)
x += x_noise
y += y_noise
return x, y, z保存ply并显示：
python
# 创建PLY结构并保存（保留原有实现）
vertices = np.zeros(len(x), dtype=[('x', 'f4'), ('y', 'f4'), ('z', 'f4')])
vertices['x'], vertices['y'], vertices['z'] = x, y, z
ply = PlyData([PlyElement.describe(vertices, 'vertex')], text=True)
ply.write('tube.ply')
# 调用可视化
visualize_ply('tube.ply')点云生成完整代码
默认半径300 fov60 测量距离2000 得到的最后效果：
python
import numpy as np
import matplotlib.pyplot as plt
from plyfile import PlyData, PlyElement
import tkinter as tk
from tkinter import ttk
import math
from tkinter import messagebox
color_map = [
(128, 0, 0), (132, 0, 0), (136, 0, 0), (140, 0, 0),
(144, 0, 0), (148, 0, 0), (152, 0, 0), (156, 0, 0),
(160, 0, 0), (164, 0, 0), (168, 0, 0), (172, 0, 0),
(176, 0, 0), (180, 0, 0), (184, 0, 0), (188, 0, 0),
(192, 0, 0), (196, 0, 0), (200, 0, 0), (204, 0, 0),
(208, 0, 0), (212, 0, 0), (216, 0, 0), (220, 0, 0),
(224, 0, 0), (228, 0, 0), (232, 0, 0), (236, 0, 0),
(240, 0, 0), (244, 0, 0), (248, 0, 0), (252, 0, 0),
(255, 0, 0), (255, 4, 0), (255, 8, 0), (255, 12, 0),
(255, 16, 0), (255, 20, 0), (255, 24, 0), (255, 28, 0),
(255, 32, 0), (255, 36, 0), (255, 40, 0), (255, 44, 0),
(255, 48, 0), (255, 52, 0), (255, 56, 0), (255, 60, 0),
(255, 64, 0), (255, 68, 0), (255, 72, 0), (255, 76, 0),
(255, 80, 0), (255, 84, 0), (255, 88, 0), (255, 92, 0),
(255, 96, 0), (255, 100, 0), (255, 104, 0), (255, 108, 0),
(255, 112, 0), (255, 116, 0), (255, 120, 0), (255, 124, 0),
(255, 128, 0), (255, 132, 0), (255, 136, 0), (255, 140, 0),
(255, 144, 0), (255, 148, 0), (255, 152, 0), (255, 156, 0),
(255, 160, 0), (255, 164, 0), (255, 168, 0), (255, 172, 0),
(255, 176, 0), (255, 180, 0), (255, 184, 0), (255, 188, 0),
(255, 192, 0), (255, 196, 0), (255, 200, 0), (255, 204, 0),
(255, 208, 0), (255, 212, 0), (255, 216, 0), (255, 220, 0),
(255, 224, 0), (255, 228, 0), (255, 232, 0), (255, 236, 0),
(255, 240, 0), (255, 244, 0), (255, 248, 0), (255, 252, 0),
(254, 255, 1), (250, 255, 6), (246, 255, 10), (242, 255, 14),
(238, 255, 18), (234, 255, 22), (230, 255, 26), (226, 255, 30),
(222, 255, 34), (218, 255, 38), (214, 255, 42), (210, 255, 46),
(206, 255, 50), (202, 255, 54), (198, 255, 58), (194, 255, 62),
(190, 255, 66), (186, 255, 70), (182, 255, 74), (178, 255, 78),
(174, 255, 82), (170, 255, 86), (166, 255, 90), (162, 255, 94),
(158, 255, 98), (154, 255, 102), (150, 255, 106), (146, 255, 110),
(142, 255, 114), (138, 255, 118), (134, 255, 122), (130, 255, 126),
(126, 255, 130), (122, 255, 134), (118, 255, 138), (114, 255, 142),
(110, 255, 146), (106, 255, 150), (102, 255, 154), (98, 255, 158),
(94, 255, 162), (90, 255, 166), (86, 255, 170), (82, 255, 174),
(78, 255, 178), (74, 255, 182), (70, 255, 186), (66, 255, 190),
(62, 255, 194), (58, 255, 198), (54, 255, 202), (50, 255, 206),
(46, 255, 210), (42, 255, 214), (38, 255, 218), (34, 255, 222),
(30, 255, 226), (26, 255, 230), (22, 255, 234), (18, 255, 238),
(14, 255, 242), (10, 255, 246), (6, 255, 250), (2, 255, 254),
(0, 252, 255), (0, 248, 255), (0, 244, 255), (0, 240, 255),
(0, 236, 255), (0, 232, 255), (0, 228, 255), (0, 224, 255),
(0, 220, 255), (0, 216, 255), (0, 212, 255), (0, 208, 255),
(0, 204, 255), (0, 200, 255), (0, 196, 255), (0, 192, 255),
(0, 188, 255), (0, 184, 255), (0, 180, 255), (0, 176, 255),
(0, 172, 255), (0, 168, 255), (0, 164, 255), (0, 160, 255),
(0, 156, 255), (0, 152, 255), (0, 148, 255), (0, 144, 255),
(0, 140, 255), (0, 136, 255), (0, 132, 255), (0, 128, 255),
(0, 124, 255), (0, 120, 255), (0, 116, 255), (0, 112, 255),
(0, 108, 255), (0, 104, 255), (0, 100, 255), (0, 96, 255),
(0, 92, 255), (0, 88, 255), (0, 84, 255), (0, 80, 255),
(0, 76, 255), (0, 72, 255), (0, 68, 255), (0, 64, 255),
(0, 60, 255), (0, 56, 255), (0, 52, 255), (0, 48, 255),
(0, 44, 255), (0, 40, 255), (0, 36, 255), (0, 32, 255),
(0, 28, 255), (0, 24, 255), (0, 20, 255), (0, 16, 255),
(0, 12, 255), (0, 8, 255), (0, 4, 255), (0, 0, 255),
(0, 0, 252), (0, 0, 248), (0, 0, 244), (0, 0, 240),
(0, 0, 236), (0, 0, 232), (0, 0, 228), (0, 0, 224),
(0, 0, 220), (0, 0, 216), (0, 0, 212), (0, 0, 208),
(0, 0, 204), (0, 0, 200), (0, 0, 196), (0, 0, 192),
(0, 0, 188), (0, 0, 184), (0, 0, 180), (0, 0, 176),
(0, 0, 172), (0, 0, 168), (0, 0, 164), (0, 0, 160),
(0, 0, 156), (0, 0, 152), (0, 0, 148), (0, 0, 144),
(0, 0, 140), (0, 0, 136), (0, 0, 132), (0, 0, 128)
]
def visualize_ply(file_path):
# 读取PLY文件
ply_data = PlyData.read(file_path)
vertices = ply_data['vertex']
# 提取坐标数据
x = vertices['x']
y = vertices['y']
z = vertices['z']
# 计算深度范围
min_depth = np.min(z)
max_depth = np.max(z)
depth_range = max_depth - min_depth
# 生成颜色映射
colors = []
for depth in z:
# 将深度映射到颜色索引 (0-255)
color_idx = int(((depth - min_depth) / depth_range) * (len(color_map)-1))
colors.append(color_map[color_idx])
# 转换为0-1范围的RGB值
colors = np.array(colors) / 255.0
# 创建3D可视化
fig = plt.figure(figsize=(10, 8))
ax = fig.add_subplot(111, projection='3d')
# 绘制点云
ax.scatter(x, y, z, c=colors, s=1, marker='o')
# 设置坐标轴标签
ax.set_xlabel('X')
ax.set_ylabel('Y')
ax.set_zlabel('Z')
# 启用交互式旋转
plt.show(block=True) # 添加block=True参数保持窗口显示
# 在文件末尾添加以下新功能
def generate_cylinder(radius, fov_deg, max_range, points=5000,noise=5,coefficient=4):
if radius <= 0 or fov_deg <= 0 or max_range <= 0:
return None
if fov_deg > 180:
return None
if radius >= max_range:
return None
if points <= 0:
return None
min_range = radius / math.sin(math.radians(fov_deg / 2))
if min_range > max_range:
return None
max_distance = math.sqrt(max_range**2 - radius**2)
min_distance = radius / math.tan(math.radians(fov_deg / 2))
if max_distance <= min_distance or max_distance <= 0 or min_distance <= 0:
return None
a_rad = math.asin(radius / max_range)
a_deg = math.degrees(a_rad)
if a_deg > fov_deg / 2:
return None
# 生成圆柱面点云
theta = np.random.uniform(0, 2*math.pi, points)
u = np.random.rand(points)
z = max_distance - (max_distance - min_distance) * np.power(u,1/coefficient)
# z = np.random.uniform(min_distance, max_distance, points)
# 转换为笛卡尔坐标（管壁表面）
radius_noise = np.random.normal(0, noise, points)
radius += radius_noise
x = radius * np.cos(theta)
y = radius * np.sin(theta)
x_noise = np.random.normal(0, noise, points)
y_noise = np.random.normal(0, noise, points)
x += x_noise
y += y_noise
return x, y, z
class CylinderGUI:
def __init__(self):
self.root = tk.Tk()
self.root.title('圆柱体内部3D点云生成器')
ttk.Label(self.root, text='管道半径:').grid(row=0, column=0)
self.radius_entry = ttk.Entry(self.root)
self.radius_entry.insert(0, '300')
self.radius_entry.grid(row=0, column=1)
# 更新输入控件
ttk.Label(self.root, text='FOV角度:').grid(row=1, column=0)
self.fov_entry = ttk.Entry(self.root)
self.fov_entry.insert(0, '60')
self.fov_entry.grid(row=1, column=1)
ttk.Label(self.root, text='测量范围:').grid(row=2, column=0)
self.range_entry = ttk.Entry(self.root)
self.range_entry.insert(0, '2000')
self.range_entry.grid(row=2, column=1)
# 生成按钮
self.generate_btn = ttk.Button(self.root, text='生成', command=self.generate)
self.generate_btn.grid(row=3, columnspan=2)
def generate(self):
try:
fov = float(self.fov_entry.get())
max_range = float(self.range_entry.get())
radius = float(self.radius_entry.get())
# 调用新的生成函数
li = generate_cylinder(
radius=radius,
fov_deg=fov,
max_range=max_range
)
if li is None:
messagebox.showwarning("警告", "未生成有效点云，请调整检测参数")
return
x, y, z = li
# 创建PLY结构并保存（保留原有实现）
vertices = np.zeros(len(x), dtype=[('x', 'f4'), ('y', 'f4'), ('z', 'f4')])
vertices['x'], vertices['y'], vertices['z'] = x, y, z
ply = PlyData([PlyElement.describe(vertices, 'vertex')], text=True)
ply.write('tube.ply')
# 调用可视化
visualize_ply('tube.ply')
except ValueError as e:
print(f"参数错误: {e}")
if __name__ == "__main__":
gui = CylinderGUI()
gui.root.mainloop()圆形3D点云检测及拟合算法
获取最密集部分点云
获取到的点云图 其在xy平面是一个圆 所以可以根据几何特性来进行拟合
由于生成的点云近处密集 远处较疏 所以先筛选出一段范围内密集的点
python
# 读取PLY文件
ply_data = PlyData.read(ply_file)
vertices = ply_data['vertex']
# 提取坐标数据
x = vertices['x']
y = vertices['y']
z = vertices['z']
# 计算深度范围
min_depth = np.min(z)
max_depth = np.max(z)
depth_range = max_depth - min_depth
# 筛选深度在指定范围内的点
x_list = []
y_list = []
z_list = []
# 动态计算bins数量（每个区间宽度为choose_depth）
num_bins = int(np.ceil(depth_range / choose_depth))
# 生成直方图
hist, bin_edges = np.histogram(z, bins=num_bins, range=(min_depth, max_depth))
max_bin_index = np.argmax(hist)
z_min = bin_edges[max_bin_index]
z_max = bin_edges[max_bin_index + 1]
x_list = []
y_list = []
z_list = []
# 筛选最密集区间内的点
for i in range(len(z)):
if z[i] >= z_min and z[i] <= z_max:
x_list.append(x[i])
y_list.append(y[i])
z_list.append(z[i])
# 在筛选点之后添加以下代码
points = np.column_stack((x_list, y_list, z_list))而后再进行圆形拟合算法
随机取三点
在这部分随机取三点 进行组合
这里的点不重复
python
def detect_circle_radom_3_points(points,retry_num=3):
circle_points=[]
circle_r=[]
for j in range(retry_num):
# 打乱点集顺序
shuffled_points = np.random.permutation(points)
i = 0
while i < len(shuffled_points) - 2:
# 取连续三个不重复的点
triplet = shuffled_points[i:i+3]
# 计算圆心和半径
center, radius = calculate_circle(*triplet)
if center is not None and radius is not None:
circle_points.append(center)
circle_r.append(radius)
# 跳过已处理的点
i += 3
return circle_points, circle_r同时 通过retry_num循环进行随机重复取值
求三点共圆进行拟合
假定圆的平面在X Y平面上 第三个点求其两个点的投影所在共圆面上 即可得到多个圆
python
def calculate_circle(p1, p2, p3):
# 转换为二维坐标（假设在XY平面）
p1_2d = p1[:2]
p2_2d = p2[:2]
p3_2d = p3[:2]
# 检查共线性
if np.linalg.matrix_rank([p2_2d - p1_2d, p3_2d - p1_2d]) < 2:
return None, None
# 构建2x2系数矩阵
A = np.array([
[2*(p2_2d[0] - p1_2d[0]), 2*(p2_2d[1] - p1_2d[1])],
[2*(p3_2d[0] - p2_2d[0]), 2*(p3_2d[1] - p2_2d[1])]
])
# 构建常数项
b = np.array([
p2_2d[0]**2 - p1_2d[0]**2 + p2_2d[1]**2 - p1_2d[1]**2,
p3_2d[0]**2 - p2_2d[0]**2 + p3_2d[1]**2 - p2_2d[1]**2
])
try:
center_2d = np.linalg.solve(A, b)
radius = np.linalg.norm(center_2d - p1_2d)
# 恢复z坐标为三点平均值
z_center = (p1[2] + p2[2] + p3[2]) / 3
return np.array([center_2d[0], center_2d[1], z_center]), radius
except np.linalg.LinAlgError:
return None, None
每次结果都是随机的
最终圆的拟合
通过过滤掉异常半径值 并进行加权平均计算 得到最后的圆心和半径：
python
def detect_circle(circle_points,circle_r):
if len(circle_points) > 0:
# 过滤异常半径（超过1.5倍IQR）
radii = np.array(circle_r)
q75, q25 = np.percentile(radii, [75, 25])
iqr = q75 - q25
mask = (radii > q25 - 1.5*iqr) & (radii < q75 + 1.5*iqr)
# 计算加权平均值（根据半径密度加权）
filtered_points = np.array(circle_points)[mask]
filtered_radii = radii[mask]
# 使用核密度估计找到最密集区域
kde = gaussian_kde(filtered_radii)
xvals = np.linspace(min(filtered_radii), max(filtered_radii), 100)
kde_vals = kde(xvals)
best_radius = xvals[np.argmax(kde_vals)]
# 找到对应半径的所有圆心
candidate_centers = filtered_points[np.abs(filtered_radii - best_radius) < iqr*0.5]
# 取三维坐标中位数作为最终圆心
final_center = np.median(candidate_centers, axis=0)
final_radius = np.median(filtered_radii)
return final_center,final_radius
如果加上刚刚的三次循环读取 则增加圆的数量：
最终结果和代码
当然 每次随机取值得到的圆也不一样 所以每次的结果都有略微偏差 可以对最后的结果进行滤波再整合
python
from tube_generate import *
import numpy as np
import matplotlib.pyplot as plt
from plyfile import PlyData, PlyElement
import math
from scipy.stats import gaussian_kde
def calculate_circle(p1, p2, p3):
# 转换为二维坐标（假设在XY平面）
p1_2d = p1[:2]
p2_2d = p2[:2]
p3_2d = p3[:2]
# 检查共线性
if np.linalg.matrix_rank([p2_2d - p1_2d, p3_2d - p1_2d]) < 2:
return None, None
# 构建2x2系数矩阵
A = np.array([
[2*(p2_2d[0] - p1_2d[0]), 2*(p2_2d[1] - p1_2d[1])],
[2*(p3_2d[0] - p2_2d[0]), 2*(p3_2d[1] - p2_2d[1])]
])
# 构建常数项
b = np.array([
p2_2d[0]**2 - p1_2d[0]**2 + p2_2d[1]**2 - p1_2d[1]**2,
p3_2d[0]**2 - p2_2d[0]**2 + p3_2d[1]**2 - p2_2d[1]**2
])
try:
center_2d = np.linalg.solve(A, b)
radius = np.linalg.norm(center_2d - p1_2d)
# 恢复z坐标为三点平均值
z_center = (p1[2] + p2[2] + p3[2]) / 3
return np.array([center_2d[0], center_2d[1], z_center]), radius
except np.linalg.LinAlgError:
return None, None
def detect_circle_radom_3_points(points,retry_num=3):
circle_points=[]
circle_r=[]
for j in range(retry_num):
# 打乱点集顺序
shuffled_points = np.random.permutation(points)
i = 0
while i < len(shuffled_points) - 2:
# 取连续三个不重复的点
triplet = shuffled_points[i:i+3]
# 计算圆心和半径
center, radius = calculate_circle(*triplet)
if center is not None and radius is not None:
circle_points.append(center)
circle_r.append(radius)
# 跳过已处理的点
i += 3
return circle_points, circle_r
def detect_circle(circle_points,circle_r):
if len(circle_points) > 0:
# 过滤异常半径（超过1.5倍IQR）
radii = np.array(circle_r)
q75, q25 = np.percentile(radii, [75, 25])
iqr = q75 - q25
mask = (radii > q25 - 1.5*iqr) & (radii < q75 + 1.5*iqr)
# 计算加权平均值（根据半径密度加权）
filtered_points = np.array(circle_points)[mask]
filtered_radii = radii[mask]
# 使用核密度估计找到最密集区域
kde = gaussian_kde(filtered_radii)
xvals = np.linspace(min(filtered_radii), max(filtered_radii), 100)
kde_vals = kde(xvals)
best_radius = xvals[np.argmax(kde_vals)]
# 找到对应半径的所有圆心
candidate_centers = filtered_points[np.abs(filtered_radii - best_radius) < iqr*0.5]
# 取三维坐标中位数作为最终圆心
final_center = np.median(candidate_centers, axis=0)
final_radius = np.median(filtered_radii)
return final_center,final_radius
def detect_tube(ply_file,choose_depth=10):
# 读取PLY文件
ply_data = PlyData.read(ply_file)
vertices = ply_data['vertex']
# 提取坐标数据
x = vertices['x']
y = vertices['y']
z = vertices['z']
# 计算深度范围
min_depth = np.min(z)
max_depth = np.max(z)
depth_range = max_depth - min_depth
# 筛选深度在指定范围内的点
x_list = []
y_list = []
z_list = []
# 动态计算bins数量（每个区间宽度为choose_depth）
num_bins = int(np.ceil(depth_range / choose_depth))
# 生成直方图
hist, bin_edges = np.histogram(z, bins=num_bins, range=(min_depth, max_depth))
max_bin_index = np.argmax(hist)
z_min = bin_edges[max_bin_index]
z_max = bin_edges[max_bin_index + 1]
x_list = []
y_list = []
z_list = []
# 筛选最密集区间内的点
for i in range(len(z)):
if z[i] >= z_min and z[i] <= z_max:
x_list.append(x[i])
y_list.append(y[i])
z_list.append(z[i])
# 在筛选点之后添加以下代码
points = np.column_stack((x_list, y_list, z_list))
circle_points = []
circle_r = []
circle_points,circle_r = detect_circle_radom_3_points(points)
circle = detect_circle(circle_points,circle_r)
colors = []
for depth in z:
# 将深度映射到颜色索引 (0-255)
color_idx = int(((depth - min_depth) / depth_range) * (len(color_map)-1))
colors.append(color_map[color_idx])
# 转换为0-1范围的RGB值
colors = np.array(colors) / 255.0
# 创建3D可视化
fig = plt.figure(figsize=(10, 8))
ax = fig.add_subplot(111, projection='3d')
# 绘制原始点云
ax.scatter(x, y, z, c=colors, s=1, marker='o', label='原始点云')
# 绘制筛选后的点云（红色突出显示）
# ax.scatter(x_list, y_list, z_list, c="black", s=1, marker='o', label='筛选点')
# 绘制筛选后的圆心和半径
# if len(circle_points) > 0:
# for center, radius in zip(circle_points, circle_r):
# # 在XY平面生成圆环
# theta = np.linspace(0, 2*np.pi, 100)
# x_circle = center[0] + radius * np.cos(theta)
# y_circle = center[1] + radius * np.sin(theta)
# # 使用圆心Z坐标的平均值作为高度
# z_circle = np.full_like(x_circle, center[2])
# ax.plot(x_circle, y_circle, z_circle, color='red', linewidth=2,
# label='检测圆环' if center is circle_points[0] else "")
print(circle)
if len(circle[0]) and circle[1]:
# 生成更密集的点用于最终圆环
theta_final = np.linspace(0, 2*np.pi, 500)
x_final = circle[0][0] + circle[1] * np.cos(theta_final)
y_final = circle[0][1] + circle[1] * np.sin(theta_final)
z_final = np.full_like(theta_final, circle[0][2])
# 使用蓝色粗线突出显示
ax.plot(x_final, y_final, z_final, color='blue', linewidth=4,
linestyle='--', label='最终检测圆环')
# 添加半径标注
ax.text(circle[0][0], circle[0][1], circle[0][2],
f'R={circle[1]:.2f}mm', color='blue', fontsize=12)
# 设置坐标轴标签
ax.set_xlabel('X')
ax.set_ylabel('Y')
ax.set_zlabel('Z')
plt.show(block=True) # 添加block=True参数保持窗口显示
if __name__ == "__main__":
detect_tube('tube.ply')优化方案
对于算法优化 可以从1/2、1/4个点中来进行取值和拟合 以减小算法开销
另外 也可以用几何规律方法、最小二乘法来进行拟合
标准差计算更新
增加了最终结果的标准差计算，使其与原始值相比较
设定为大于50则不认定拟合成功 并画出黑色的圆
python
radius_std = np.std(filtered_radii - final_radius)如果拟合成功 则还是蓝色的圆：
附录：列表的赋值类型和py打包
列表赋值
BUG复现
闲来无事写了个小程序 代码如下：
python
# -*- coding: utf-8 -*-
"""
Created on Fri Nov 19 19:47:01 2021
@author: 16016
"""
a_list = ['0','1','2','3','4','5','6','7','8','9','10','11','12','13','14','15']
#print(len(a_list))
#b_list = ['','','','','','','','','','','','','','','','']
c_list = [[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[]]
#for i in range(16):
if len(a_list):
for j in range(16):
a_list[j]=str(a_list[j])+'_'+str(j)
print("序号:",j)
print('a_list:\n',a_list)
c_list[j]=a_list
print('c_list[0]:\n',c_list[0])
print('\n')
# b_list[j]=a_list[7],a_list[8]
# print(b_list[j])
# 写入到Excel：
#print(c_list,'\n') 我在程序中 做了一个16次的for循环 把列表a的每个值后面依次加上"_"和循环序号
比如循环第x次 就是把第x位加上_x 这一位变成x_x 我在输出测试中 列表a的每一次输出也是对的
循环16次后列表a应该变成['0_0', '1_1', '2_2', '3_3', '4_4', '5_5', '6_6', '7_7', '8_8', '9_9', '10_10', '11_11', '12_12', '13_13', '14_14', '15_15'] 这也是对的
同时 我将每一次循环时列表a的值 写入到空列表c中 比如第x次循环 就是把更改以后的列表a的值 写入到列表c的第x位
第0次循环后 c[0]的值应该是['0_0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '10', '11', '12', '13', '14', '15'] 这也是对的
但是在第1次循环以后 c[0]的值就一直在变 变成了c[x]的值
相当于把c_list[0]变成了c_list[1]...以此类推 最后得出的列表c的值也是每一项完全一样
我不明白这是怎么回事
我的c[0]只在第0次循环时被赋值了 但是后面它的值跟着在改变
如图：
第一次老出bug 赋值以后 每次循环都改变c[0]的值 搞了半天都没搞出来
无论是用appen函数添加 还是用二维数组定义 或者增加第三个空数组来过渡 都无法解决
代码改进
后来在我华科同学的指导下 突然想到赋值可以赋的是个地址 地址里面的值一直变化 导致赋值也一直变化 于是用第二张图的循环套循环深度复制实现了
代码如下：
python
# -*- coding: utf-8 -*-
"""
Created on Fri Nov 19 19:47:01 2021
@author: 16016
"""
a_list = ['0','1','2','3','4','5','6','7','8','9','10','11','12','13','14','15']
#print(len(a_list))
#b_list = ['','','','','','','','','','','','','','','','']
c_list = [[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[]]
#for i in range(16):
if len(a_list):
for j in range(16):
a_list[j]=str(a_list[j])+'_'+str(j)
print("序号:",j)
print('a_list:\n',a_list)
for i in range(16):
c_list[j].append(a_list[i])
print('c_list[0]:\n',c_list[0])
print('\n')
# b_list[j]=a_list[7],a_list[8]
# print(b_list[j])
# 写入到Excel：
print(c_list,'\n') 解决了问题
优化
第三次是请教了老师 用copy函数来赋真值
代码如下：
python
# -*- coding: utf-8 -*-
"""
Created on Fri Nov 19 19:47:01 2021
@author: 16016
"""
a_list = ['0','1','2','3','4','5','6','7','8','9','10','11','12','13','14','15']
#print(len(a_list))
#b_list = ['','','','','','','','','','','','','','','','']
c_list = [[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[]]
#for i in range(16):
if len(a_list):
for j in range(16):
a_list[j]=str(a_list[j])+'_'+str(j)
print("序号:",j)
print('a_list:\n',a_list)
c_list[j]=a_list.copy()
print('c_list[0]:\n',c_list[0])
print('\n')
# b_list[j]=a_list[7],a_list[8]
# print(b_list[j])
# 写入到Excel：
#print(c_list,'\n') 同样能解决问题
最后得出问题 就是指针惹的祸！
a_list指向的是个地址 而不是值 a_list[i]指向的才是单个的值 copy()函数也是复制值而不是地址
如果这个用C语言来写 就直观一些了 难怪C语言是基础 光学Python不学C 遇到这样的问题就解决不了
C语言yyds Python是什么垃圾弱智语言
总结
由于Python无法单独定义一个值为指针或者独立的值 所以只能用列表来传送
只要赋值是指向一个列表整体的 那么就是指向的一个指针内存地址 解决方法只有一个 那就是将每个值深度复制赋值（子列表内的元素提取出来重新依次连接） 或者用copy函数单独赋值
如图测试：
部分代码：
python
# -*- coding: utf-8 -*-
"""
Created on Sat Nov 20 16:45:48 2021
@author: 16016
"""
def text1():
A=[1,2,3]
B=[[],[],[]]
for i in range(len(A)):
A[i]=A[i]+i
B[i]=A
print(B)
def text2():
A=[1,2,3]
B=[[],[],[]]
A[0]=A[0]+0
B[0]=A
print(B)
A[1]=A[1]+1
B[1]=A
print(B)
A[2]=A[2]+2
B[2]=A
print(B)
if __name__ == '__main__':
text1()
print('\n')
text2()py打包
Pyinstaller打包exe（包括打包资源文件 绝不出错版）
依赖包及其对应的版本号
PyQt5 5.10.1
PyQt5-Qt5 5.15.2
PyQt5-sip 12.9.0
pyinstaller 4.5.1
pyinstaller-hooks-contrib 2021.3
Pyinstaller -F setup.py 打包exe
Pyinstaller -F -w setup.py 不带控制台的打包
Pyinstaller -F -i xx.ico setup.py 打包指定exe图标打包
打包exe参数说明：
-F：打包后只生成单个exe格式文件；
-D：默认选项，创建一个目录，包含exe文件以及大量依赖文件；
-c：默认选项，使用控制台(就是类似cmd的黑框)；
-w：不使用控制台；
-p：添加搜索路径，让其找到对应的库；
-i：改变生成程序的icon图标。
如果要打包资源文件
则需要对代码中的路径进行转换处理
另外要注意的是 如果要打包资源文件 则py程序里面的路径要从./xxx/yy换成xxx/yy 并且进行路径转换
但如果不打包资源文件的话 最好路径还是用作./xxx/yy 并且不进行路径转换
python
def get_resource_path(relative_path):
if hasattr(sys, '_MEIPASS'):
return os.path.join(sys._MEIPASS, relative_path)
return os.path.join(os.path.abspath("."), relative_path)而后再spec文件中的datas部分加入目录
如：
python
a = Analysis(['cxk.py'],
pathex=['D:\\Python Test\\cxk'],
binaries=[],
datas=[('root','root')],
hiddenimports=[],
hookspath=[],
hooksconfig={},
runtime_hooks=[],
excludes=[],
win_no_prefer_redirects=False,
win_private_assemblies=False,
cipher=block_cipher,
noarchive=False)而后直接Pyinstaller -F setup.spec即可
如果打包的文件过大则更改spec文件中的excludes 把不需要的库写进去（但是已经在环境中安装了的）就行
这些不要了的库在上一次编译时的shell里面输出
比如：
然后用pyinstaller --clean -F 某某.spec