有了openpose数据，下一步如何分析人体姿态
0
收藏
小白问：用paddlehub 的openpose_body_estimation得到数据后，下一步如何做可以判断到一个图片里的人有没有类似举手、低头的行为呢？
import paddlehub as hub
#image_test = "test1.jpg"
image_test = "test2.jpg"
model = hub.Module(name='openpose_body_estimation')
result = model.predict(image_test,visualization=True)
mylog = open('recode.log', mode = 'a',encoding='utf-8')
print(result, file=mylog)
mylog.close()
处理的图片结果图片:
记录的返回信息：
{'candidate': array([[2.56000000e+02, 5.50000000e+01, 9.26660120e-01, 0.00000000e+00],
[1.37000000e+02, 5.80000000e+01, 9.97184694e-01, 1.00000000e+00],
[4.09000000e+02, 6.30000000e+01, 9.93659139e-01, 2.00000000e+00],
[2.49000000e+02, 1.07000000e+02, 9.61833715e-01, 3.00000000e+00],
[1.22000000e+02, 1.10000000e+02, 9.54079151e-01, 4.00000000e+00],
[4.15000000e+02, 1.16000000e+02, 9.18871701e-01, 5.00000000e+00],
[8.00000000e+01, 1.07000000e+02, 8.92051935e-01, 6.00000000e+00],
[2.01000000e+02, 1.10000000e+02, 8.85539711e-01, 7.00000000e+00],
[3.77000000e+02, 1.20000000e+02, 9.49104965e-01, 8.00000000e+00],
[6.40000000e+01, 1.84000000e+02, 9.53616440e-01, 9.00000000e+00],
[3.76000000e+02, 1.87000000e+02, 8.54201436e-01, 1.00000000e+01],
[1.80000000e+02, 1.88000000e+02, 9.29694414e-01, 1.10000000e+01],
[3.81000000e+02, 2.03000000e+02, 8.28347564e-01, 1.20000000e+01],
[2.14000000e+02, 2.04000000e+02, 8.52818012e-01, 1.30000000e+01],
[1.06000000e+02, 2.06000000e+02, 4.58440721e-01, 1.40000000e+01],
[2.95000000e+02, 1.04000000e+02, 9.23401177e-01, 1.50000000e+01],
[4.54000000e+02, 1.11000000e+02, 8.89868736e-01, 1.60000000e+01],
[1.65000000e+02, 1.15000000e+02, 8.67759407e-01, 1.70000000e+01],
[3.22000000e+02, 1.70000000e+02, 9.11746562e-01, 1.80000000e+01],
[1.64000000e+02, 1.84000000e+02, 6.69153214e-01, 1.90000000e+01],
[4.60000000e+02, 1.88000000e+02, 8.61127496e-01, 2.00000000e+01],
[1.49000000e+02, 1.89000000e+02, 6.78965807e-01, 2.10000000e+01],
[3.07000000e+02, 2.08000000e+02, 8.75384748e-01, 2.20000000e+01],
[4.33000000e+02, 2.25000000e+02, 8.05526435e-01, 2.30000000e+01],
[8.10000000e+01, 2.20000000e+02, 7.37039268e-01, 2.40000000e+01],
[3.95000000e+02, 2.25000000e+02, 7.32784748e-01, 2.50000000e+01],
[2.35000000e+02, 2.30000000e+02, 7.25677431e-01, 2.60000000e+01],
[1.12000000e+02, 2.21000000e+02, 8.85800719e-01, 2.70000000e+01],
[2.01000000e+02, 2.23000000e+02, 9.09173369e-01, 2.80000000e+01],
[3.74000000e+02, 2.42000000e+02, 7.71196306e-01, 2.90000000e+01],
[9.70000000e+01, 3.29000000e+02, 6.43031716e-01, 3.00000000e+01],
[2.44000000e+02, 3.29000000e+02, 6.26661301e-01, 3.10000000e+01],
[3.77000000e+02, 3.29000000e+02, 1.36328757e-01, 3.20000000e+01],
[1.37000000e+02, 2.19000000e+02, 6.17978990e-01, 3.30000000e+01],
[2.90000000e+02, 2.24000000e+02, 7.25328445e-01, 3.40000000e+01],
[4.44000000e+02, 2.27000000e+02, 7.78226972e-01, 3.50000000e+01],
[3.38000000e+02, 2.21000000e+02, 9.42319214e-01, 3.60000000e+01],
[1.52000000e+02, 2.23000000e+02, 8.77975941e-01, 3.70000000e+01],
[4.07000000e+02, 2.67000000e+02, 9.35000598e-01, 3.80000000e+01],
[1.57000000e+02, 3.29000000e+02, 7.72364140e-01, 3.90000000e+01],
[3.07000000e+02, 3.29000000e+02, 4.35196221e-01, 4.00000000e+01],
[4.25000000e+02, 3.29000000e+02, 1.65057734e-01, 4.10000000e+01],
[2.48000000e+02, 4.60000000e+01, 9.59142804e-01, 4.20000000e+01],
[1.30000000e+02, 4.80000000e+01, 9.49559450e-01, 4.30000000e+01],
[4.01000000e+02, 5.50000000e+01, 9.16016042e-01, 4.40000000e+01],
[2.65000000e+02, 4.80000000e+01, 9.81730044e-01, 4.50000000e+01],
[1.47000000e+02, 5.10000000e+01, 9.93008912e-01, 4.60000000e+01],
[4.19000000e+02, 5.50000000e+01, 9.18187320e-01, 4.70000000e+01],
[2.28000000e+02, 4.70000000e+01, 9.46349442e-01, 4.80000000e+01],
[1.14000000e+02, 5.00000000e+01, 8.41706038e-01, 4.90000000e+01],
[3.91000000e+02, 6.10000000e+01, 9.13940072e-01, 5.00000000e+01],
[2.73000000e+02, 5.00000000e+01, 3.11306089e-01, 5.10000000e+01],
[1.55000000e+02, 5.90000000e+01, 6.00095391e-01, 5.20000000e+01],
[4.30000000e+02, 6.10000000e+01, 8.75765264e-01, 5.30000000e+01]]), 'subset': array([[ 1. , 4. , 6. , 9. , 14. ,
17. , 19. , 21. , 24. , 27. ,
30. , 33. , 37. , 39. , 43. ,
46. , 49. , 52. , 30.03926984, 18. ],
[ 0. , 3. , 7. , 11. , 13. ,
15. , 18. , 22. , 26. , 28. ,
31. , 34. , 36. , 40. , 42. ,
45. , 48. , 51. , 30.76957117, 18. ],
[ 2. , 5. , 8. , 10. , 12. ,
16. , 20. , 23. , 25. , 29. ,
32. , 35. , 38. , 41. , 44. ,
47. , 50. , 53. , 30.85516463, 18. ]]), 'data': array([[[ 37, 51, 33],
[ 35, 49, 31],
[ 32, 48, 30],
...,
[ 16, 48, 24],
[ 16, 48, 24],
[ 15, 47, 23]],
[[ 33, 47, 29],
[ 32, 46, 28],
[ 30, 46, 28],
...,
[ 17, 49, 25],
[ 17, 49, 25],
[ 16, 48, 24]],
[[ 26, 40, 22],
[ 27, 41, 23],
[ 26, 42, 24],
...,
[ 19, 51, 27],
[ 18, 50, 26],
[ 18, 50, 26]],
...,
[[ 34, 33, 23],
[ 59, 39, 28],
[ 76, 56, 39],
...,
[151, 66, 234],
[166, 85, 242],
[175, 103, 247]],
[[ 31, 37, 20],
[ 56, 39, 20],
[ 90, 66, 38],
...,
[174, 95, 246],
[166, 87, 238],
[162, 83, 234]],
[[ 38, 44, 27],
[ 61, 44, 25],
[ 93, 69, 41],
...,
[174, 95, 246],
[167, 88, 239],
[162, 83, 234]]], dtype=uint8)}
{'candidate': array([[ 90. , 38. , 0.87060368, 0. ],
[102. , 65. , 0.77316552, 1. ],
[ 70. , 67. , 0.71633345, 2. ],
[ 26. , 87. , 0.79980302, 3. ],
[ 44. , 45. , 0.77932775, 4. ],
[135. , 62. , 0.65731257, 5. ],
[178. , 94. , 0.80417168, 6. ],
[162. , 48. , 0.7487663 , 7. ],
[ 82. , 32. , 0.88247955, 8. ],
[ 96. , 30. , 0.95693362, 9. ],
[ 75. , 38. , 0.76395959, 10. ],
[109. , 32. , 0.89137608, 11. ]]), 'subset': array([[ 0. , 1. , 2. , 3. , 4. ,
5. , 6. , 7. , -1. , -1. ,
-1. , -1. , -1. , -1. , 8. ,
9. , 10. , 11. , 19.70410003, 12. ]]), 'data': array([[[255, 255, 255],
[255, 255, 255],
[255, 255, 255],
...,
[209, 218, 205],
[208, 217, 204],
[207, 216, 203]],
[[255, 255, 255],
[255, 255, 255],
[255, 255, 255],
...,
[209, 218, 205],
[208, 217, 204],
[207, 216, 203]],
[[255, 255, 255],
[255, 255, 255],
[255, 255, 255],
...,
[209, 218, 205],
[208, 217, 204],
[207, 216, 203]],
...,
[[240, 242, 236],
[240, 242, 236],
[240, 242, 236],
...,
[190, 201, 181],
[190, 201, 181],
[189, 200, 180]],
[[240, 242, 236],
[240, 242, 236],
[240, 242, 236],
...,
[189, 200, 180],
[189, 200, 180],
[189, 200, 180]],
[[241, 243, 237],
[241, 243, 237],
[241, 243, 237],
...,
[188, 199, 179],
[188, 199, 179],
[188, 199, 179]]], dtype=uint8)}
ps:是不是需要找训练集打标签训练，用coco数据集可以吗，一时不知如何下手，Thanks♪(･ω･)ﾉ
收藏
点赞
0
个赞
请登录后评论
TOP
切换版块
同学，解决了吗，遇到了一样的问题
https://xasgkfp.diytrade.com/
http://xasgkfp.diytrade.com/
http://xakfp.simplesite.com/
http://cdfapiao.simplesite.com/
https://xasgkfp.diytrade.com
http://xasgkfp.diytrade.com