一、OpenCV
OpenCV（开源计算机视觉库）是一个用于计算机视觉和机器学习的开源库，适用于处理图片和视频数据。
1. 图像读取和显示:
<code>
import cv2
# 读取图像
img = cv2.imread('test.jpg',1)
# 显示图像
cv2.imshow('image',img)
cv2.waitKey(0)
cv2.destroyAllWindows()
</code>
2. 边缘检测:
<code>
import cv2
# 读取图像
img = cv2.imread('test.jpg',1)
# 灰度化
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
# 边缘检测
edges = cv2.Canny(gray,50,150)
# 显示边缘
cv2.imshow('image',edges)
cv2.waitKey(0)
cv2.destroyAllWindows()
</code>
3. 视频捕获:
<code>
import cv2
cap = cv2.VideoCapture(0)
while(True):
# 逐帧捕获
ret, frame = cap.read()
# 显示帧
cv2.imshow('frame',frame)
# 退出
if cv2.waitKey(1) & 0xFF == ord('q'):
break
# 释放摄像头
cap.release()
cv2.destroyAllWindows()
</code>
二、Open3D
Open3D 是一个现代化的3D数据处理库，以 Python 和 C++ 两种编程语言扩展，旨在专注实现高效、可扩展、易用的3D数据处理和机器学习算法。
1. 加载点云数据:
<code>
import open3d as o3d
# 读入点云
pcd = o3d.io.read_point_cloud("test.ply")
# 可视化
o3d.visualization.draw_geometries([pcd])
</code>
2. 配准:
<code>
import open3d as o3d
# 读入两个点云
source = o3d.io.read_point_cloud("source.ply")
target = o3d.io.read_point_cloud("target.ply")
# 配准
trans_init = np.asarray([[1,0,0,0],[0,1,0,0],[0,0,1,0],[0,0,0,1]])
reg_p2p = o3d.registration.registration_icp(
source, target, max_correspondence_distance_coarse, trans_init,
o3d.registration.TransformationEstimationPointToPoint(),
o3d.registration.ICPConvergenceCriteria(relative_fitness=1e-6,
relative_rmse=1e-6, max_iteration=1000))
# 刚体变换
source.transform(reg_p2p.transformation)
# 可视化
o3d.visualization.draw_geometries([source, target])
</code>
三、OpenAI
OpenAI 是一个致力于人工智能的非营利研究公司，旨在建立人工智能 “安全”的范例。
1. 文本生成:
<code>
import openai
openai.api_key = "YOUR_API_KEY"
prompt = "In a shocking turn of events, scientists have discovered a herd of unicorns living in a remote, previously unexplored valley, in the Andes Mountains. Even more surprising to the researchers was the fact that the unicorns spoke perfect English."
model = "text-davinci-002"
response = openai.Completion.create(engine=model, prompt=prompt, max_tokens=60)
print(response.choices[0].text)
</code>
2. 图像分类:
<code>
import openai
openai.api_key = "YOUR_API_KEY"
model_id = "image-alpha-001"
input_url = "https://images.pexels.com/photos/4499335/pexels-photo-4499335.jpeg"
output_format = "json"
response = openai.Image.create(model_id, input_url, output_format)
print(response.output_url)
</code>
以上是Python开发者必备了解的几个Open库，它们可以帮助开发者快速从事计算机视觉、3D数据处理和机器学习等方向的工作，减少重复开发和提高效率。
原创文章，作者：小蓝，如若转载，请注明出处：https://www.506064.com/n/227202.html
微信扫一扫 
支付宝扫一扫 