def calculate_orb_values(img,resize_ratio,resize_method,crop_amount):
# crop if crop_amount above 0
if crop_amount>0:
img = image_managment.crop_image(img, crop_amount)
# resize if resize ratio greater than 0
if resize_ratio>0:
img = image_managment.resize_image(img, ratio=resize_ratio, method=resize_method)
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY) # convert image to gray-scale
orb = cv2.ORB() # initialise our ORB detector
kp,des = orb.detectAndCompute(gray,None) # detect keypoints and compute descriptors
return kp,des,gray
def calculate_sift_values(img,resize_ratio,resize_method,crop_amount):
# crop if crop_amount above 0
if crop_amount>0:
img = image_managment.crop_image(img, crop_amount)
# resize if resize ratio greater than 0
if resize_ratio>0:
img = image_managment.resize_image(img, ratio=resize_ratio, method=resize_method)
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY) # convert image to gray-scale
sift = cv2.SIFT() # initialise our SIFT detector
#sift.nfeatures = n_features # set the number of feature to retain
kp, des = sift.detectAndCompute(gray,None) # detect keypoints and compute descriptors
return kp,des,gray
def calculate_surf_values(img,hessian_threshold,resize_ratio,resize_method,crop_amount):
# crop if crop_amount above 0
if crop_amount>0:
img = image_managment.crop_image(img, crop_amount)
# resize if resize ratio greater than 0
if resize_ratio>0:
img = image_managment.resize_image(img, ratio=resize_ratio, method=resize_method)
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY) # convert image to gray-scale
surf = cv2.SURF() # initialise our surf detector
surf.hessianThreshold = hessian_threshold # set the hessian threshold for our surf detector
surf.extended = True # use the 128 dimension descriptors
kp, des = surf.detectAndCompute(gray,None) # detect keypoints and compute descriptors
return kp,des,gray