def main():
# create camera object. initial calibration will be executed by constructor
cap = ht301_hacklib.HT301()
ret, frame = cap.read() # read frame from thermal camera
time.sleep(3)
ret, frame = cap.read() # read frame from thermal camera
offset = frame.astype(np.float32)
np.save("noise_pattern_calibration.npy", offset)
print("Calibration successfull!")
cap.release()
return 0
#!/usr/bin/python3
import numpy as np
import cv2
import math
import ht301_hacklib
import utils
import time
draw_temp = True
cap = ht301_hacklib.HT301()
cv2.namedWindow("HT301", cv2.WINDOW_NORMAL)
while (True):
ret, frame = cap.read()
info, lut = cap.info()
frame = frame.astype(np.float32)
# Sketchy auto-exposure
frame -= frame.min()
frame /= frame.max()
frame = (np.clip(frame, 0, 1) * 255).astype(np.uint8)
frame = cv2.applyColorMap(frame, cv2.COLORMAP_JET)
if draw_temp:
utils.drawTemperature(frame, info['Tmin_point'], info['Tmin_C'],
(55, 0, 0))
utils.drawTemperature(frame, info['Tmax_point'], info['Tmax_C'],
(0, 0, 85))
utils.drawTemperature(frame, info['Tcenter_point'], info['Tcenter_C'],
def main():
########### camera #############
# create camera object. initial calibration will be executed by constructor
cap = ht301_hacklib.HT301()
# read calibration array from file
offset = np.load("noise_pattern_calibration.npy")
# create a Contrast Limited Adaptive Histogram Equalization object. default: 5.0, (6, 6)
clahe = cv2.createCLAHE(clipLimit=5.0, tileGridSize=(6, 6))
# open v4l2 output device
videooutput = os.open(VIDEO_OUT, os.O_RDWR)
# configure parameters for output device
vid_format = v4l2_format()
vid_format.type = V4L2_BUF_TYPE_VIDEO_OUTPUT
if fcntl.ioctl(videooutput, VIDIOC_G_FMT, vid_format) < 0:
print("ERROR: unable to get video format!")
return -1
framesize = VID_WIDTH * VID_HEIGHT * 3
vid_format.fmt.pix.width = VID_WIDTH
vid_format.fmt.pix.height = VID_HEIGHT
# pixel format
vid_format.fmt.pix.pixelformat = V4L2_PIX_FMT_RGB24
vid_format.fmt.pix.sizeimage = framesize
vid_format.fmt.pix.field = V4L2_FIELD_NONE
if fcntl.ioctl(videooutput, VIDIOC_S_FMT, vid_format) < 0:
print("ERROR: unable to set video format!")
return -1
################################
# video loop
while (True):
# cycle only one map per button press
if rc_channel > 1800:
ch_state = True
if prev_ch_state == False:
cyclecolormaps()
else:
ch_state = False
prev_ch_state = ch_state
ret, frame = cap.read() # read frame from thermal camera
info, lut = cap.info(
) # get hottest and coldest spot and its temperatures
# automatic gain control
frame = frame.astype(np.float32)
if calibration_offset == True: # apply calibration offset
frame = frame - offset + np.mean(offset)
frame = (255 *
((frame - frame.min()) / (frame.max() - frame.min()))).astype(
np.uint8) # cast frame to values from 0 to 255
if colormaps[selectedmap] == -1:
# digital detail enhancement algorithm
framebuffer = frame
clahe.apply(framebuffer, frame)
frame = cv2.applyColorMap(frame, colormaps[selectedmap - 1])
else:
# apply colormap
frame = cv2.applyColorMap(frame, colormaps[selectedmap])
if flipped_camera == True: # rotate the temperature points with the image if the thermal camera is mounted upside down
(coldx, coldy) = info['Tmin_point']
(warmx, warmy) = info['Tmax_point']
coldestpoint = (VID_WIDTH - coldx, VID_HEIGHT - coldy)
warmestpoint = (VID_WIDTH - warmx, VID_HEIGHT - warmy)
frame = cv2.flip(frame, -1) # flip the frame
else:
coldestpoint = info['Tmin_point']
warmestpoint = info['Tmax_point']
if draw_temp: # color: BGR
utils.drawTemperature(frame, coldestpoint, info['Tmin_C'],
(200, 200, 200)) # coldest spot
utils.drawTemperature(frame, warmestpoint, info['Tmax_C'],
(30, 30, 30)) # hottest spot
#utils.drawTemperature(frame, info['Tcenter_point'], info['Tcenter_C'], (0, 255, 255)) # center spot
# output
frame = cv2.cvtColor(
frame, cv2.COLOR_BGR2RGB
) # convert opencv bgr to rgb for the v4l2 pixelformat
written = os.write(videooutput,
frame.data) # write frame to output device
cap.release()
if mavlink:
vehicle.close()
return 0
def main():
########### camera #############
cap = ht301_hacklib.HT301(
) # create camera object. initial calibration will be executed by contructor
videooutput = os.open(VIDEO_OUT, os.O_RDWR) # open v4l2 output device
# configure params for output device
vid_format = v4l2_format()
vid_format.type = V4L2_BUF_TYPE_VIDEO_OUTPUT
if fcntl.ioctl(videooutput, VIDIOC_G_FMT, vid_format) < 0:
print("ERROR: unable to get video format!")
return -1
framesize = VID_WIDTH * VID_HEIGHT * 3
vid_format.fmt.pix.width = VID_WIDTH
vid_format.fmt.pix.height = VID_HEIGHT
# NOTE: change this according to below filters...
# Chose one from the supported formats on Chrome: YUV420, Y16, Z16, INVZ,
# YUYV, RGB24, MJPEG, JPEG
vid_format.fmt.pix.pixelformat = V4L2_PIX_FMT_RGB24
vid_format.fmt.pix.sizeimage = framesize
vid_format.fmt.pix.field = V4L2_FIELD_NONE
if fcntl.ioctl(videooutput, VIDIOC_S_FMT, vid_format) < 0:
print("ERROR: unable to set video format!")
return -1
################################
while (True):
# cycle only one map per button press
##if vehicle.channels[rc_channel] > 1800:
## ch_state = True
## if prev_ch_state == False:
## cyclecolormaps()
##else:
## ch_state = False
##prev_ch_state = ch_state
ret, frame = cap.read() # read frame from thermal camera
info, lut = cap.info(
) # get hottest and coldest spot and its temperatures
frame = frame.astype(np.float32)
if selectedmap != 0: # do not apply anything when 0 is selected. original frame
# Sketchy auto-exposure
frame -= frame.min()
frame /= frame.max()
frame = (np.clip(frame, 0, 1) * 255).astype(np.uint8)
frame = cv2.applyColorMap(frame,
colormaps[selectedmap]) # apply colormap
# rotate the temperature points with the image
(coldx, coldy) = info['Tmin_point']
(warmx, warmy) = info['Tmax_point']
coldestpoint = (VID_WIDTH - coldx, VID_HEIGHT - coldy)
warmestpoint = (VID_WIDTH - warmx, VID_HEIGHT - warmy)
frame = cv2.flip(frame, -1)
if draw_temp:
utils.drawTemperature(frame, coldestpoint, info['Tmin_C'],
(0, 0, 255)) # coldest spot
utils.drawTemperature(frame, warmestpoint, info['Tmax_C'],
(255, 0, 0)) # hottest spot
#utils.drawTemperature(frame, info['Tcenter_point'], info['Tcenter_C'], (0,255,255)) # center spot
# write frame to output device
written = os.write(videooutput, frame.data)
cap.release()
##vehicle.close()
return 0