def computeHoFromImage(starList):
# the leading theory is that this approximation will only work when the stars are closeby
# that is, the small angle approximation works
pixScale = 24.1 # 5 # arcsec/pixel
pixDegrees = pixScale / 3600
K = 3600 / 24.2
imageWidth = 3872 # pix
imageHeight = 2592 # pix
xc = imageWidth / 2
yc = imageHeight / 2
naturalBias = -4.9399 # -11.0701857994#-5.25 #calc from lyra_oct9 basic optimum
# naturalBias =-4.7598 #this is teh TRUE
# naturalBias =-3.3920
# naturalBias = -4.9749
# naturalBias = -5.2764
# PITCH CALIBRATION SUPER IMPORTANT
pitch = 62.62 # 39.76#62.62 #deg
# i'll need to get the correct roll value from the IMU
roll = -0.54 # 1.9 #-0.32-1.9056-3.19+4444#-0.5 # -2.4#-0.54
rollOffset = -4.7598 # basic optimal point
# rollOffset = 0.4523 #this one is the TRUE
# rollOffset = 10.4774
# rollOffset = -5.2764
# strip 'annotations'
# starList = starList['annotations']
shortList = [] # make a short list of the stars we're really interested in
for star in starList:
# if the star has an alternate name its probably big or bright or both
# if len(star['names'])>1:
# pixel offsets
star["newx"] = star["pixelx"] * ST.cosd(roll + rollOffset) + star["pixely"] * ST.sind(roll + rollOffset)
star["newy"] = -star["pixelx"] * ST.sind(roll + rollOffset) + star["pixely"] * ST.cosd(roll + rollOffset)
star["xOff"] = star["pixelx"] - xc
star["yOff"] = yc - star["pixely"]
# rotation will go here
# compute offset from roll angle
# i think the order we do the roll and pitch operations is kind of important
# star['ho'] = pitch + star['yOff']*pixDegrees + naturalBias # i feel like its plus and the matlab is actually whats wrong
star["ho"] = -pixDegrees * star["xOff"] * ST.sind(roll + rollOffset) + (
pixDegrees * star["yOff"] + (pitch + naturalBias)
) * ST.cosd(roll + rollOffset)
# shortList.append(star)
print star
# print "\n\n\n shortList \n\n\n" %shortList
return [starList, pitch, roll]
def calcD(Li, Bi, Lf, Bf):
d = 60*m.sqrt((Li-Lf)**2*ST.cosd(Bf) + (Bi - Bf)**2)
return d
