Drone-Images-Mosaicing
Drone-Images-Mosaicing copied to clipboard
adityajain07
Later version having issues impelementing your code
could you update your code for later vesions of python am interested in some ideas here
Hi Martin,
Can you please specify the python version and also post the error snapshot?
this is the code `import cv2 import numpy as np from matplotlib import pyplot as plt import glob import time import os
import math
FilepathFirst = os.path.abspath("/Users/uSER/Downloads/Documents/DJI_0122.JPG") FilepathFirst1 = os.path.abspath("/Users/uSER/Downloads/Documents/DJI_0124.JPG") FilepathAll = os.path.abspath("/Users/uSER/Downloads/Documents/*.JPG") img1 = cv2.imread(FilepathFirst) #img2 = cv2.imread(FilepathFirst1) tic = time.process_time() test = []
def warpImages(img1, img2, H): rows1, cols1 = img1.shape[:2] rows2, cols2 = img2.shape[:2]
list_of_points_1 = np.float32([[0,0], [0,rows1], [cols1,rows1], [cols1,0]]).reshape(-1,1,2)
temp_points = np.float32([[0,0], [0,rows2], [cols2,rows2], [cols2,0]]).reshape(-1,1,2)
list_of_points_2 = cv2.perspectiveTransform(temp_points, H)
list_of_points = np.concatenate((list_of_points_1, list_of_points_2), axis=0)
[x_min, y_min] = np.int32(list_of_points.min(axis=0).ravel() - 0.5)
[x_max, y_max] = np.int32(list_of_points.max(axis=0).ravel() + 0.5)
translation_dist = [-x_min,-y_min]
H_translation = np.array([[1, 0, translation_dist[0]], [0, 1, translation_dist[1]], [0,0,1]])
print ("Min and Max:", x_min, y_min, x_max, y_max)
print ("Translation Distance:", translation_dist )
output_img = cv2.warpPerspective(img2, H_translation.dot(H), (x_max-x_min, y_max-y_min))
FrameSize = output_img.shape
NewImage = img2.shape
output_img[translation_dist[1]:rows1+translation_dist[1], translation_dist[0]:cols1+translation_dist[0]] = img1
OriginR = int(list_of_points_2[0][0][1])
OriginC = int(list_of_points_2[0][0][0])
# if the origin of projected image is out of bounds, then mapping to ()
if OriginR < 0:
OriginR = 0
if OriginC < 0:
OriginC = 0
# Clipping the new image, if it's size is more than the frame
if NewImage[0] > FrameSize[0]-OriginR:
img2 = img2[0:FrameSize[0]-OriginR,:]
if NewImage[1] > FrameSize[1]-OriginC:
img2 = img2[:,0:FrameSize[1]-OriginC]
print ("Image 2 Magic size:", img2.shape)
output_img[OriginR:NewImage[0]+OriginR, OriginC:NewImage[1]+OriginC] = img2
return output_img
images = sorted(glob.glob(FilepathAll)) # for reading images n = 10000; # no of features to extract
def giveMosaic(FirstImage, no):
EList = [] # this stores the average reprojection error
ImgList = [] # No of images stitched
Matches = [] # this stores the number of good matches at every stage
i = 1
heightM,widthM = FirstImage.shape[:2]
FirstImage = cv2.resize(FirstImage, (widthM / 4, heightM / 4))
RecMosaic = FirstImage
for name in images[1:]:
print (name)
image = cv2.imread(name)
sift = cv2.SIFT(no)
######## Resize them (they are too big)
# Get their dimensions
height,width = image.shape[:2]
print heightM, widthM, height, width
image = cv2.resize(image, (width / 4, height / 4))
###########################
# Find the features
kp1= des1 = sift.detectAndCompute(RecMosaic,None) # kp are the keypoints, des are the descriptors
kp2= des2 = sift.detectAndCompute(image,None)
########### FLANN Matcher ##########
FLANN_INDEX_KDTREE = 0
index_params = dict(algorithm = FLANN_INDEX_KDTREE, trees = 5)
search_params = dict(checks = 50)
flann = cv2.FlannBasedMatcher(index_params, search_params)
matches = flann.knnMatch(des1,des2,k=2)
#############
# store all the good matches as per Lowe's ratio test.
good = []
allPoints = []
for m,n in matches:
if m.distance < 0.7*n.distance:
good.append(m)
allPoints.append(m)
Matches.append(len(good))
print ("Good_Matches:", len(good))
##################################
#### Finding the homography #########
src_pts = np.float32([ kp1[m.queryIdx].pt for m in good ]).reshape(-1,1,2)
dst_pts = np.float32([ kp2[m.trainIdx].pt for m in good ]).reshape(-1,1,2)
all_src_pts = np.float32([ kp1[m.queryIdx].pt for m in allPoints ]).reshape(-1,1,2)
all_dst_pts = np.float32([ kp2[m.trainIdx].pt for m in allPoints ]).reshape(-1,1,2)
M, mask = cv2.findHomography(dst_pts, src_pts, cv2.RANSAC,6.0)
###################################
#### Finding the euclidean distance error ####
list1 = np.array(src_pts)
list2 = np.array(dst_pts)
list2 = np.reshape(list2, (len(list2), 2))
ones = np.ones(len(list1))
TestPoints = np.transpose(np.reshape(list1, (len(list1), 2)))
print ("Length:", np.shape(TestPoints), np.shape(ones))
TestPointsHom = np.vstack((TestPoints, ones))
print ("Homogenous Points:", np.shape(TestPointsHom))
projectedPointsH = np.matmul(M, TestPointsHom) # projecting the points in test image to collage image using homography matrix
projectedPointsNH = np.transpose(np.array([np.true_divide(projectedPointsH[0,:], projectedPointsH[2,:]), np.true_divide(projectedPointsH[1,:], projectedPointsH[2,:])]))
print ("list2 shape:", np.shape(list2))
print ("NH Points shape:", np.shape(projectedPointsNH))
print ("Raw Error Vector:", np.shape(np.linalg.norm(projectedPointsNH-list2, axis=1)))
Error = int(np.sum(np.linalg.norm(projectedPointsNH-list2, axis=1)))
print ("Total Error:", Error)
AvgError = np.divide(np.array(Error), np.array(len(list1)))
print ("Average Error:", AvgError)
##################
i+=1
RecMosaic = warpImages(RecMosaic, image, M)
cv2.imwrite("FinalMosaicTemp.jpg", RecMosaic)
print (i)
EList.append(AvgError)
ImgList.append(i)
if i==40:
break
cv2.imwrite("FinalMosaic.jpg", RecMosaic)
return EList, ImgList, Matches
ErrorList,ImgNumbers,GoodMatches = giveMosaic(img1, 10000) toc = time.process_time() print(toc-tic) ` Bellow is the error that am getting
https://github.com/martinchez/Mosaicking-DJI-Aerial-imagery-using-python this is the repository to the images that am using if you could help
You have to downgrade your cv2 version and the script is written in python 2.x so you have to change the code also. hope this will help you.
How do I change the code.
On Mon, Dec 14, 2020, 17:19 Deepak Kumar Saxena [email protected] wrote:
You have to downgrade your cv2 version and the script is written in python 2.x so you have to change the code also. hope this will help you.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/adityajain07/Drone-Images-Mosaicing/issues/1#issuecomment-744470983, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHTTBCA2XTSBE2R45OYHIADSUYNHLANCNFSM4UOQKYNQ .
!pip install opencv-python==3.4.2.17 !pip install opencv-contrib-python==3.4.2.17
3.4.9.33, 3.4.10.35, 3.4.10.37, 3.4.11.39, 3.4.11.41, 3.4.11.43, 3.4.11.45, 4.1.2.30, 4.2.0.32, 4.2.0.34, 4.3.0.36, 4.3.0.3
8, 4.4.0.40, 4.4.0.42, 4.4.0.44, 4.4.0.46)
ERROR: No matching distribution found for opencv-python==3.4.2.17```
please go through my repo of multiple image-stitching and stitching a pair of images here is the link https://github.com/deepzsenu/Image-stitching-multiple-and-a-pair-of-images-