Load a ground truth file from /poses, and get a N x 3 x 4 data cube:



In [4]:

    
cd ../../src/opencv_vision_challenge/









    



[Error 3] The system cannot find the path specified: u'../../src/opencv_vision_challenge/'
c:\code\rick\opencv_vision_challenge



In [5]:

    
import numpy as np
import matplotlib as mpl
from matplotlib import pyplot as plt
%matplotlib inline



In [6]:

    
a = np.loadtxt('00.txt')



In [7]:

    
a.shape









    Out[7]:





(4541L, 12L)



In [8]:

    
poses = a.reshape(4541, 3, 4)



In [9]:

    
poses.shape









    Out[9]:





(4541L, 3L, 4L)

Extract just the translation vectors from pose matrices:



In [10]:

    
T = poses[:, :, 3].copy()
T.shape









    Out[10]:





(4541L, 3L)

Extract individual coords from the translation vectors:



In [11]:

    
X = T[:, 0]
Y = T[:, 1]
Z = T[:, 2]



In [12]:

    
X.shape









    Out[12]:





(4541L,)

Plot a 'map' of the data (note: z is forward, x is to the right of the car):



In [13]:

    
plt.plot(X, Z)
plt.show()



In [13]: