The Collector¶

This is a simple example of how to use the library to read a data file and plot some stuff.

In [2]:

%matplotlib inline 
from __future__ import division
import cv2         # opencv itself
import numpy as np # matrix manipulations
import random
from math import sqrt

from matplotlib import pyplot as plt
import pylab
pylab.rcParams['figure.figsize'] = (6.0, 6.0)

from the_collector import BagReader

In [3]:

def normalize(sensor):
    for i, s in enumerate(sensor):
        x,y,z = s
        mag = sqrt(x**2 + y**2 + z**2)
        sensor[i] = [x/mag, y/mag, z/mag]
    return sensor

In [4]:

%ls

imu-1-2.json*             sensor_data.mp4*          simple-calibration.ipynb*
imu-1.json*               sensors.ipynb*

In [5]:

bgr = BagReader()
bgr.use_compression = True
data = bgr.load('imu-1-2.json')

In [7]:

for key in data.keys():
    print(key, len(data[key]))

gyro 585
camera 585
accel 585
mag 585
b64keys 1

In [8]:

gyros = []
gdt = []
start = data['gyro'][0][1]
for g, dt in data['gyro']:
    gyros.append(g)
    gdt.append(dt - start)
plt.plot(gdt, gyros)
plt.title('Gyros')
plt.xlabel('Time [sec]')
plt.ylabel('Rads/sec')
plt.grid(True)
plt.legend(['x','y','z'])

Out[8]:

<matplotlib.legend.Legend at 0x119a863c8>

In [9]:

mag = []
mdt = []
start = data['mag'][0][1]
for m, dt in data['mag']:
    mag.append(m)
    mdt.append(dt - start)
    
# mag = normalize(mag)
    
plt.plot(mdt, mag)
plt.title('mag')
plt.xlabel('Time [sec]')
plt.ylabel('Gauss')
plt.grid(True)
plt.legend(['x','y','z'])

Out[9]:

<matplotlib.legend.Legend at 0x119b2c4a8>

In [10]:

accel = []
adt = []
start = data['accel'][0][1]
for a, dt in data['accel']:
    accel.append(a)
    adt.append(dt - start)
    
accel = normalize(accel)
    
plt.plot(adt, accel)
plt.title('Accel')
plt.xlabel('Time [sec]')
plt.ylabel('Gs')
plt.grid(True)
plt.legend(['x','y','z'])

Out[10]:

<matplotlib.legend.Legend at 0x119cf36a0>

In [12]:

f,ts = data['camera'][100]
plt.imshow(f)
print(f.shape);

(240, 320, 3)

In [13]:

# simple function to save a video
import platform
def videoWrite(frames, fname='out.mp4'):
    frame_height, frame_width, _ = frames[0].shape
    
    # pick a good encoder for the current OS
    sys = platform.system()
    if sys in ['Darwin']:
        fourcc = 'avc1'
    else:
        fourcc = 'mjpg'
        
    out = cv2.VideoWriter(
        fname,
        cv2.VideoWriter_fourcc(*fourcc), 
        30, 
        (frame_width,frame_height))
    for frame in frames:
        out.write(frame)
    out.release()

In [14]:

# save camera frames to a video
frames = []
for f, ts in data['camera']:
    frames.append(f)

videoWrite(frames, 'sensor_data.mp4')

In [ ]: