EasyMocap/code/mytools/reconstruction.py
2021-01-25 19:37:23 +08:00

106 lines
3.5 KiB
Python

'''
* @ Date: 2020-09-14 11:01:52
* @ Author: Qing Shuai
@ LastEditors: Qing Shuai
@ LastEditTime: 2021-01-25 16:06:41
@ FilePath: /EasyMocap/code/mytools/reconstruction.py
'''
import numpy as np
def solveZ(A):
u, s, v = np.linalg.svd(A)
X = v[-1, :]
X = X / X[3]
return X[:3]
def projectN3(kpts3d, Pall):
# kpts3d: (N, 3)
nViews = len(Pall)
kp3d = np.hstack((kpts3d[:, :3], np.ones((kpts3d.shape[0], 1))))
kp2ds = []
for nv in range(nViews):
kp2d = Pall[nv] @ kp3d.T
kp2d[:2, :] /= kp2d[2:, :]
kp2ds.append(kp2d.T[None, :, :])
kp2ds = np.vstack(kp2ds)
return kp2ds
def simple_reprojection_error(kpts1, kpts1_proj):
# (N, 3)
error = np.mean((kpts1[:, :2] - kpts1_proj[:, :2])**2)
return error
def simple_triangulate(kpts, Pall):
# kpts: (nViews, 3)
# Pall: (nViews, 3, 4)
# return: kpts3d(3,), conf: float
nViews = len(kpts)
A = np.zeros((nViews*2, 4), dtype=np.float)
result = np.zeros(4)
result[3] = kpts[:, 2].sum()/(kpts[:, 2]>0).sum()
for i in range(nViews):
P = Pall[i]
A[i*2, :] = kpts[i, 2]*(kpts[i, 0]*P[2:3,:] - P[0:1,:])
A[i*2 + 1, :] = kpts[i, 2]*(kpts[i, 1]*P[2:3,:] - P[1:2,:])
result[:3] = solveZ(A)
return result
def simple_recon_person(keypoints_use, Puse, config=None, ret_repro=False):
eps = 0.01
nJoints = keypoints_use[0].shape[0]
if isinstance(keypoints_use, list):
keypoints_use = np.stack(keypoints_use)
out = np.zeros((nJoints, 4))
for nj in range(nJoints):
keypoints = keypoints_use[:, nj]
if (keypoints[:, 2] > 0.01).sum() < 2:
continue
out[nj] = simple_triangulate(keypoints, Puse)
if config is not None:
# remove the false limb with the help of limb
for (i, j), mean_std in config['skeleton'].items():
ii, jj = min(i, j), max(i, j)
if out[ii, -1] < eps:
out[jj, -1] = 0
if out[jj, -1] < eps:
continue
length = np.linalg.norm(out[ii, :3] - out[jj, :3])
if abs(length - mean_std['mean'])/(3*mean_std['std']) > 1:
# print((i, j), length, mean_std)
out[jj, :] = 0
# 计算重投影误差
kpts_repro = projectN3(out, Puse)
square_diff = (keypoints_use[:, :, :2] - kpts_repro[:, :, :2])**2
conf = np.repeat(out[None, :, -1:], len(Puse), 0)
kpts_repro = np.concatenate((kpts_repro, conf), axis=2)
if conf.sum() < 3: # 至少得有3个有效的关节
repro_error = 1e3
else:
conf2d = conf *(keypoints_use[:, :, -1:] > 0.01)
# (nViews, nJoints): reprojection error for each joint in each view
repro_error_joint = np.sqrt(square_diff.sum(axis=2, keepdims=True))*conf2d
# remove the not valid joints
# remove the bad views
repro_error = repro_error_joint.sum()/conf.sum()
if ret_repro:
return out, repro_error, kpts_repro
return out, repro_error
def check_limb(keypoints3d, limb_means, thres=0.5):
# keypoints3d: (nJ, 4)
valid = True
cnt = 0
for (src, dst), val in limb_means.items():
if not (keypoints3d[src, 3] > 0 and keypoints3d[dst, 3] > 0):
continue
cnt += 1
# 计算骨长
l_est = np.linalg.norm(keypoints3d[src, :3] - keypoints3d[dst, :3])
if abs(l_est - val['mean'])/val['mean']/val['std'] > thres:
valid = False
break
# 至少两段骨头可以使用
valid = valid and cnt > 2
return valid