Mind+Python+Mediapipe项目——AI健身之开合跳
本帖最后由 云天 于 2021-12-3 23:02 编辑【项目背景】
要说燃脂运动，相信很多人首先会想到“波比跳”，当然啦对于很多新手来说做波比跳并不是特别的容易，所以他们就会选择一种相对比较简单但效果同样很不错的——开合跳！长时间锻炼，需要毅力坚持。但也可以让枯燥的运动，变的有乐趣。今天这个项目让人工智能陪我们一起快乐健身。
【项目设计】
使用Mind+Python模式下加载Google的开源Mediapipe人工智能算法库，识别人体姿态，利用动作中两臂与躯体的夹角及两腿夹角的变化来判断开合跳次数，并通过Pinpong库控制LED灯实时显示次数。
其中使用到了Python文件操作功能，边制作边学习。
【测试程序】
使用网络视频进行合跳测试，标注关键信息点。
import numpy as np
import time
import cv2
import PoseModule as pm
cap = cv2.VideoCapture('hetiao4.mp4')
detector = pm.poseDetector()
count = 0
dir = 0
pTime = 0
success=True
while success:
success, img = cap.read()
if success:
img = cv2.resize(img, (640, 480))
img = detector.findPose(img, False)
lmList = detector.findPosition(img, False)
if len(lmList) != 0:
# 双腿
angle1 = detector.findAngle(img, 28, 23, 27)
# 双臂
angle2 = detector.findAngle(img, 15, 11, 23)
angle3 = detector.findAngle(img, 16, 12, 24)
print(angle1,angle2,angle3)
cTime = time.time()
fps = 1 / (cTime - pTime)
pTime = cTime
cv2.putText(img, str(int(fps)), (50, 100), cv2.FONT_HERSHEY_PLAIN, 5,(255, 0, 0), 5)
cv2.imshow("Image", img)
cv2.waitKey(1)
cap.release()
cv2.destroyAllWindows()
https://www.bilibili.com/video/BV1XY411s7Du?share_source=copy_web
【获取标准数据】
通过对标准动作检测，获取动作标准数据，并存储在数据文件中。
import numpy as np
import time
import cv2
import PoseModule as pm
import os,sys
cap = cv2.VideoCapture('ht.mp4')
detector = pm.poseDetector()
count = 0
dir = 0
pTime = 0
success=True
angle1=[]
angle2=[]
angle3=[]
def max_min(a):
h = []
l = []
for i in range(1, len(a)-1):
if(a < a and a < a):
h.append(a)
elif(a > a and a > a):
l.append(a)
if(len(h) == 0):
h.append(max(a))
if(len(l) == 0):
l.append(min(a))
print(int(np.mean(h)),int(np.mean(l)))
return(int(np.mean(h)),int(np.mean(l)))
while success:
success, img = cap.read()
if success:
img = cv2.resize(img, (640, 480))
img = detector.findPose(img, False)
lmList = detector.findPosition(img, False)
if len(lmList) != 0:
# 双腿
angle1.append( detector.findAngle(img, 28, 23, 27))
# 双臂
angle2.append( detector.findAngle(img, 15, 11, 23))
angle3.append( detector.findAngle(img, 16, 12, 24))
cTime = time.time()
fps = 1 / (cTime - pTime)
pTime = cTime
cv2.putText(img, str(int(fps)), (50, 100), cv2.FONT_HERSHEY_PLAIN, 5,(255, 0, 0), 5)
cv2.imshow("Image", img)
cv2.waitKey(1)
a1,b1=max_min(angle1)
a2,b2=max_min(angle2)
a3,b3=max_min(angle3)
fo = open("data.txt", "w")
fo.write(str(a1)+","+str(b1)+","+str(a2)+","+str(b2)+","+str(a3)+","+str(b3))
# 关闭文件
fo.close()
cap.release()
cv2.destroyAllWindows()
调用函数计算出两臂、两腿夹角的最大角度和最小角度，并通过文件操作将数据写入“data.txt”文件中。
a1,b1=max_min(angle1)
a2,b2=max_min(angle2)
a3,b3=max_min(angle3)
fo = open("data.txt", "w")
fo.write(str(a1)+","+str(b1)+","+str(a2)+","+str(b2)+","+str(a3)+","+str(b3))
【开合跳计数程序】
1、读取标准数据文件，并将数据转换为列表
from pathlib import Path
import sys
my_file = Path("data.txt")
if not my_file.is_file():
print("没有标准数据文件")
sys.exit()
else:
f = open("data.txt",encoding = "utf-8")
data=f.read()
data=data.split(",")
f.close()
2、利用标准数据对动作进行判定，是否完成一次完整动作，并在屏幕显示
# 双腿夹角
angle1 = detector.findAngle(img, 28, 23, 27)
# 双臂与躯干夹角
angle2 = detector.findAngle(img, 15, 11, 23)
angle3 = detector.findAngle(img, 16, 12, 24)
#print(angle1,angle2,angle3)
per1 = np.interp(angle1, (int(data), int(data)), (0, 100))
per2 = np.interp(angle2, (int(data), int(data)), (100, 0))
per3 = np.interp(angle3, (int(data), int(data)), (0, 100))
light = int(np.interp(angle1, (220, 310), (119, 0)))
# print(angle, per)
# 计算个数
per=per1+per2+per3
#print(per1,per2,per3)
if per == 300:
if dir == 0:
count += 0.5
dir = 1
if per == 0:
if dir == 1:
count += 0.5
dir = 0
#print(count)
cv2.putText(img, str(int(count)), (45, 460), cv2.FONT_HERSHEY_PLAIN, 7,(255, 0, 0), 8)
https://www.bilibili.com/video/BV1ES4y1X7Vt?share_source=copy_web
【LED炫灯显示计数】
1、Pinpong库初始化from pinpong.board import Board,Pin,NeoPixel
NEOPIXEL_PIN = Pin.P0
PIXELS_NUM = 120 #灯数
Board("microbit").begin()#初始化
npX = NeoPixel(Pin(NEOPIXEL_PIN), PIXELS_NUM)2、亮灯显示计数
light = int(np.interp(int(count)), (0, 120), (0, 120)))
npX.rainbow(0,light,0,0x0000FF)【完整程序】
import numpy as np
import time
import cv2
import PoseModule as pm
from pathlib import Path
import sys
my_file = Path("data.txt")
if not my_file.is_file():
print("没有标准数据文件")
sys.exit()
else:
f = open("data.txt",encoding = "utf-8")
data=f.read()
data=data.split(",")
f.close()
from pinpong.board import Board,Pin,NeoPixel
NEOPIXEL_PIN = Pin.P0
PIXELS_NUM = 120 #灯数
Board("microbit").begin()#初始化
npX = NeoPixel(Pin(NEOPIXEL_PIN), PIXELS_NUM)
cap = cv2.VideoCapture('d.mp4')
cap.set(3, 640)
cap.set(4, 480)
detector = pm.poseDetector()
count = 0
dir = 0
pTime = 0
success=True
while success:
success, img = cap.read()
if success:
img = cv2.resize(img, (640, 480))
img = detector.findPose(img, False)
lmList = detector.findPosition(img, False)
if len(lmList) != 0:
# 双腿夹角
angle1 = detector.findAngle(img, 28, 23, 27)
# 双臂与躯干夹角
angle2 = detector.findAngle(img, 15, 11, 23)
angle3 = detector.findAngle(img, 16, 12, 24)
#print(angle1,angle2,angle3)
per1 = np.interp(angle1, (int(data), int(data)), (0, 100))
per2 = np.interp(angle2, (int(data), int(data)), (100, 0))
per3 = np.interp(angle3, (int(data), int(data)), (0, 100))
# print(angle, per)
# 计算个数
per=per1+per2+per3
#print(per1,per2,per3)
if per == 300:
if dir == 0:
count += 0.5
dir = 1
if per == 0:
if dir == 1:
count += 0.5
dir = 0
#print(count)
light = int(np.interp(int(count)), (0, 120), (0, 120)))
npX.rainbow(0,light,0,0x0000FF)
cv2.putText(img, str(int(count)), (45, 460), cv2.FONT_HERSHEY_PLAIN, 7,(255, 0, 0), 8)
cTime = time.time()
fps = 1 / (cTime - pTime)
pTime = cTime
cv2.putText(img, str(int(fps)), (50, 100), cv2.FONT_HERSHEY_PLAIN, 5,(255, 0, 0), 5)
cv2.imshow("Image", img)
cv2.waitKey(1)
cap.release()
cv2.destroyAllWindows()
【演示视频】
https://www.bilibili.com/video/BV1bS4y1X7HS?share_source=copy_web
【PoseModule.py】
以上文件所引用的“PoseModule.py”文件。
import math
import mediapipe as mp
import cv2
class poseDetector():
def __init__(self, mode=False, upBody=False, smooth=True,
detectionCon=0.85, trackCon=0.5):
self.mode = mode
self.upBody = upBody
self.smooth = smooth
self.detectionCon = detectionCon
self.trackCon = trackCon
self.mpDraw = mp.solutions.drawing_utils
self.mpPose = mp.solutions.pose
self.pose = self.mpPose.Pose(self.mode, self.upBody, self.smooth,
self.detectionCon, self.trackCon)
def findPose(self, img, draw=True):
imgRGB = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
self.results = self.pose.process(imgRGB)
if self.results.pose_landmarks:
if draw:
self.mpDraw.draw_landmarks(img, self.results.pose_landmarks,
self.mpPose.POSE_CONNECTIONS)
return img
def findPosition(self, img, draw=True):
self.lmList = []
if self.results.pose_landmarks:
for id, lm in enumerate(self.results.pose_landmarks.landmark):
h, w, c = img.shape
# print(id, lm)
cx, cy = int(lm.x * w), int(lm.y * h)
self.lmList.append()
if draw:
cv2.circle(img, (cx, cy), 5, (255, 0, 0), cv2.FILLED)
return self.lmList
def midpoint(self,img,p1,p2,draw=True):
x1, y1 = self.lmList
x2, y2 = self.lmList
x3=int((x1+x2)/2)
y3=int((y1+y2)/2)
if draw:
cv2.circle(img, (x3, y3), 10, (0, 0, 255), cv2.FILLED)
cv2.circle(img, (x3, y3), 15, (0, 0, 255), 2)
point={"x":x3,"y":y3}
return point
def findAngle(self, img, p1, p2, p3, draw=True):
# Get the landmarks
x1, y1 = self.lmList
x2, y2 = self.lmList
x3, y3 = self.lmList
# Calculate the Angle
angle = math.degrees(math.atan2(y3 - y2, x3 - x2) -
math.atan2(y1 - y2, x1 - x2))
if angle < 0:
angle += 360
# print(angle)
# Draw
if draw:
cv2.line(img, (x1, y1), (x2, y2), (255, 255, 255), 3)
cv2.line(img, (x3, y3), (x2, y2), (255, 255, 255), 3)
cv2.circle(img, (x1, y1), 10, (0, 0, 255), cv2.FILLED)
cv2.circle(img, (x1, y1), 15, (0, 0, 255), 2)
cv2.circle(img, (x2, y2), 10, (0, 0, 255), cv2.FILLED)
cv2.circle(img, (x2, y2), 15, (0, 0, 255), 2)
cv2.circle(img, (x3, y3), 10, (0, 0, 255), cv2.FILLED)
cv2.circle(img, (x3, y3), 15, (0, 0, 255), 2)
cv2.putText(img, str(int(angle)), (x2 - 50, y2 + 50),
cv2.FONT_HERSHEY_PLAIN, 2, (0, 0, 255), 2)
return angle
想问下， if per == 300: 这个300是什么意思的，我看了好久看不懂呢
页:
[1]