Develop Reference

Cropping in OpenCV return images rotated 90 degrees

I want to crop images according to their right frame. I have about 10000 of hand X-ray images to preprocess, and what I have done so far:
Apply Gaussian Blur and Threshold (Binary + Otsu) on the image.
Apply dilation to get a single object (in this case a hand).
Used cv2.findContours() to draw outline along the edges around the hand.
Used cv2.boundingRect() to find the right frame, and then cv2.minAreaRect() and cv2.boxPoints to get the right points for the bounding box.
Used cv2.warpPerspective to adjust image according to height and width.
The code below describes the above:
import os
import cv2
import numpy as np
from matplotlib import pyplot as plt
# Load image, create mask, grayscale, Gaussian blur, Otsu's threshold
img_path = "sample_image.png"
image = cv2.imread(image_path)
original = image.copy()
blank = np.zeros(image.shape[:2], dtype = np.uint8)
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
blur = cv2.GaussianBlur(gray, (33,33), 0)
thresh = cv2.threshold(blur, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)[1]
# Merge text into a single contour
kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (5,5))
dilate = cv2.dilate(thresh, kernel, iterations = 3)
# Find contours
cnts = cv2.findContours(dilate, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
cnts = cnts[0] if len(cnts) == 2 else cnts[1]
cnts = sorted(cnts, key = lambda x: cv2.boundingRect(x)[0])
for c in cnts:
# Filter using contour area and aspect ratio (x1 = width, y1 = height)
x, y, x1, y1 = cv2.boundingRect(c)
if (x1 > 500) and (y1 > 700):
rect = cv2.minAreaRect(c)
box = cv2.boxPoints(rect)
box = np.int0(box)
width = int(rect[1][0])
height = int(rect[1][1])
src_pts = box.astype("float32")
dst_pts = np.array([[0, height-1], [0, 0],
[width-1, 0], [width-1, height-1]], dtype="float32")
M = cv2.getPerspectiveTransform(src_pts, dst_pts)
warped = cv2.warpPerspective(image, M, (width, height))
plt.imshow(warped)
If you have a look at some of the images in the folder, those are the inputs. When I run these images through the code above, I get an output like this. Some of them are cropped nicely (straightened), however, some of them are cropped with 90 degree rotations. Is there a code to counter the 'rotating 90 degrees output' problem?
Here are some images:
Image Inputs: Four X-ray examples
Image Outputs: Returns images that are 90 degrees rotated
Image Outputs wanted: Straightened image (Just used Photoshop to straighten them. Dont want to do this for 10000 images...)
UPDATE:
I edited the code according to below-mentioned suggestions. After running the some samples, it now returns images that are now 90 degrees slanted to the right.
Input images:
Output images:
I doubt it's because of the quality of the images. Maybe it's got to do with OpenCV's minAreaRect()? or boxPoints?
FINAL UPDATE:
According to #Prashant Maurya, the code was updated with a function added to detect whether the position of the hand is left or right. And then mapping src_pts to right dst_pts. Full code is shown below.

Hi there are two changes which will correct the output:
The width and height taken in the code is in the wrong order ie: width: 1470 & height: 1118 just switch the values:
Map src_pts to right dst_pts the current code is mapping top left
corner to bottom left therefore the image is being rotated.
Added function to detect whether image is right tilted or left and rotate and rotate it accordingly
Full code with changes is:
import os
import cv2
import numpy as np
from matplotlib import pyplot as plt
# Load image, create mask, grayscale, Gaussian blur, Otsu's threshold
img_path = "xray1.png"
image = cv2.imread(img_path)
cv2.imshow("image original", image)
cv2.waitKey(10000)
original = image.copy()
blank = np.zeros(image.shape[:2], dtype = np.uint8)
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
blur = cv2.GaussianBlur(gray, (33,33), 0)
thresh = cv2.threshold(blur, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)[1]
# Merge text into a single contour
kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (5,5))
dilate = cv2.dilate(thresh, kernel, iterations = 3)
# Find contours
cnts = cv2.findContours(dilate, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
cnts = cnts[0] if len(cnts) == 2 else cnts[1]
cnts = sorted(cnts, key = lambda x: cv2.boundingRect(x)[0])
def get_tilt(box):
tilt = "Left"
x_list = [coord[0] for coord in box]
y_list = [coord[1] for coord in box]
print(x_list)
print(y_list)
x_list = sorted(x_list)
y_list = sorted(y_list)
print(x_list)
print(y_list)
for coord in box:
if coord[0] == x_list[0]:
index = y_list.index(coord[1])
print("Index: ", index)
if index == 1:
tilt = "Left"
else:
tilt = "Right"
return tilt
for c in cnts:
# Filter using contour area and aspect ratio (x1 = width, y1 = height)
x, y, x1, y1 = cv2.boundingRect(c)
if (x1 > 500) and (y1 > 700):
rect = cv2.minAreaRect(c)
print("rect",rect)
box = cv2.boxPoints(rect)
box = np.int0(box)
# print("rect:", box)
tilt = get_tilt(box)
src_pts = box.astype("float32")
if tilt == "Left":
width = int(rect[1][1])
height = int(rect[1][0])
dst_pts = np.array([[0, 0],
[width-1, 0], [width-1, height-1], [0, height-1]], dtype="float32")
else:
width = int(rect[1][0])
height = int(rect[1][1])
dst_pts = np.array([[0, height-1], [0, 0],
[width-1, 0], [width-1, height-1]], dtype="float32")
print("Src pts:", src_pts)
print("Dst pts:", dst_pts)
M = cv2.getPerspectiveTransform(src_pts, dst_pts)
warped = cv2.warpPerspective(image, M, (width, height))
print("Showing image ..")
# plt.imshow(warped)
cv2.imshow("image crop", warped)
cv2.waitKey(10000)

Particle detection with Python OpenCV

I'm looking for a proper solution how to count particles and measure their sizes in this image:
In the end I have to obtain the lists of particles' coordinates and area squares. After some search on the internet I realized there are 3 approaches for particles detection:
blobs
Contours
connectedComponentsWithStats
Looking at different projects I assembled some code with the mix of it.
import pylab
import cv2
import numpy as np
Gaussian blurring and thresholding
original_image = cv2.imread(img_path)
img = original_image
img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
img = cv2.GaussianBlur(img, (5, 5), 0)
img = cv2.blur(img, (5, 5))
img = cv2.medianBlur(img, 5)
img = cv2.bilateralFilter(img, 6, 50, 50)
max_value = 255
adaptive_method = cv2.ADAPTIVE_THRESH_GAUSSIAN_C
threshold_type = cv2.THRESH_BINARY
block_size = 11
img_thresholded = cv2.adaptiveThreshold(img, max_value, adaptive_method, threshold_type, block_size, -3)
filter small objects
min_size = 4
nb_components, output, stats, centroids = cv2.connectedComponentsWithStats(img, connectivity=8)
sizes = stats[1:, -1]
nb_components = nb_components - 1
# for every component in the image, you keep it only if it's above min_size
for i in range(0, nb_components):
if sizes[i] < min_size:
img[output == i + 1] = 0
generation of Contours for filling holes and measurements. pos_list and size_list is what we were looking for
contours, hierarchy = cv2.findContours(img, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
pos_list = []
size_list = []
for i in range(len(contours)):
area = cv2.contourArea(contours[i])
size_list.append(area)
(x, y), radius = cv2.minEnclosingCircle(contours[i])
pos_list.append((int(x), int(y)))
for the self-check, if we plot these coordinates over the original image
pts = np.array(pos_list)
pylab.figure(0)
pylab.imshow(original_image)
pylab.scatter(pts[:, 0], pts[:, 1], marker="x", color="green", s=5, linewidths=1)
pylab.show()
We might get something like the following:
And... I'm not really satisfied with the results. Some clearly visible particles are not included, on the other side, some doubt fluctuations of intensity have been counted. I'm playing now with different filters' settings, but the feeling is it's wrong.
If someone knows how to improve my solution, please share.

Since the particles are in white and the background in black, we can use Kmeans Color Quantization to segment the image into two groups with cluster=2. This will allow us to easily distinguish between particles and the background. Since the particles may be very tiny, we should try to avoid blurring, dilating, or any morphological operations which may alter the particle contours. Here's an approach:
Kmeans color quantization. We perform Kmeans with two clusters, grayscale, then Otsu's threshold to obtain a binary image.
Filter out super tiny noise. Next we find contours, remove tiny specs of noise using contour area filtering, and collect each particle (x, y) coordinate and its area. We remove tiny particles on the binary mask by "filling in" these contours to effectively erase them.
Apply mask onto original image. Now we bitwise-and the filtered mask onto the original image to highlight the particle clusters.
Kmeans with clusters=2
Result
Number of particles: 204
Average particle size: 30.537
Code
import cv2
import numpy as np
import pylab
# Kmeans
def kmeans_color_quantization(image, clusters=8, rounds=1):
h, w = image.shape[:2]
samples = np.zeros([h*w,3], dtype=np.float32)
count = 0
for x in range(h):
for y in range(w):
samples[count] = image[x][y]
count += 1
compactness, labels, centers = cv2.kmeans(samples,
clusters,
None,
(cv2.TERM_CRITERIA_EPS + cv2.TERM_CRITERIA_MAX_ITER, 10000, 0.0001),
rounds,
cv2.KMEANS_RANDOM_CENTERS)
centers = np.uint8(centers)
res = centers[labels.flatten()]
return res.reshape((image.shape))
# Load image
image = cv2.imread('1.png')
original = image.copy()
# Perform kmeans color segmentation, grayscale, Otsu's threshold
kmeans = kmeans_color_quantization(image, clusters=2)
gray = cv2.cvtColor(kmeans, cv2.COLOR_BGR2GRAY)
thresh = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)[1]
# Find contours, remove tiny specs using contour area filtering, gather points
points_list = []
size_list = []
cnts, _ = cv2.findContours(thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)[-2:]
AREA_THRESHOLD = 2
for c in cnts:
area = cv2.contourArea(c)
if area < AREA_THRESHOLD:
cv2.drawContours(thresh, [c], -1, 0, -1)
else:
(x, y), radius = cv2.minEnclosingCircle(c)
points_list.append((int(x), int(y)))
size_list.append(area)
# Apply mask onto original image
result = cv2.bitwise_and(original, original, mask=thresh)
result[thresh==255] = (36,255,12)
# Overlay on original
original[thresh==255] = (36,255,12)
print("Number of particles: {}".format(len(points_list)))
print("Average particle size: {:.3f}".format(sum(size_list)/len(size_list)))
# Display
cv2.imshow('kmeans', kmeans)
cv2.imshow('original', original)
cv2.imshow('thresh', thresh)
cv2.imshow('result', result)
cv2.waitKey()

How to remove Edge Text in image using Opencv

I want to remove the text on the edged in the image
I have used the following code but it does not work it also remove the text in the center
Input:
Output:
import cv2
import matplotlib.pyplot as plt
import glob
import os
import numpy as np
def crop_buttom_text(img):
""" Remove the text from the bottom edge of img """
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
#blur = cv2.GaussianBlur(gray, (9,9), 0) # No need for blurring
thresh = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)[1]
# Create rectangular structuring element and dilate
kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (30, 1)) # Use horizontal line as kernel - dilate horizontally.
dilate = cv2.dilate(thresh, kernel, iterations=1)
#kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (10,10))
#dilate = cv2.morphologyEx(dilate, cv2.MORPH_OPEN, kernel) # No need for opening
# Find contours and draw rectangle
cnts = cv2.findContours(dilate, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)[-2] # [-2] indexing takes return value before last (due to OpenCV compatibility issues).
#cnts = cnts[0] if len(cnts) == 2 else cnts[1] # [-2] is shorter....
res_img = img.copy() # Copy img to res_img - in case there is no edges text.
for c in cnts:
x, y, w, h = cv2.boundingRect(c)
y2 = y + h # Bottom y coordinate of the bounding rectangle
if (y2 >= img.shape[0]):
# If the rectangle touches the bottom of the img
res_img = res_img[0:y-1, :].copy() # Crop rows from first row to row y-1
return res_img
def remove_lines(image_path,outdir):
image = cv2.imread(image_path)
img1 = crop_buttom_text(image)
img2 = crop_buttom_text(np.rot90(img1)) # Rotate by 90 degrees and crop.
img3 = crop_buttom_text(np.rot90(img2)) # Rotate by 90 degrees and crop.
img4 = crop_buttom_text(np.rot90(img3)) # Rotate by 90 degrees and crop.
output_img = np.rot90(img4)
cv2.imwrite(os.path.join(outdir,os.path.basename(image_path)), output_img)
for jpgfile in glob.glob(r'/content/Dataset/*'):
print(jpgfile)
remove_lines(jpgfile,r'/content/output')
How can i modify the above code to remove the text around the edge

How do you remove the text which is at the edge? Here is a rough way to attack the problem: remove every connected component which touches the border.
To solve this, you can add a 1 pixel border, extract the connected components and then use the one corresponding to the border as a binary mask.
The mask becomes
The result is
Notice how a 7 and few commas get removed as well.
from cv2 import cv2
import numpy as np
# Load original image
img = cv2.imread('kShDc.jpg', cv2.IMREAD_GRAYSCALE)
# Save original dimensions
h, w = img.shape[:2]
# Ensure only bilevel image with white as foreground
_, bimg = cv2.threshold(img, 127, 255, cv2.THRESH_BINARY_INV)
# Add one pixel border
bimg = cv2.copyMakeBorder(bimg, 1, 1, 1, 1, cv2.BORDER_CONSTANT, None, 255)
# Extract connected components (Spaghetti is the fastest algorithm)
nlabels, label_image = cv2.connectedComponentsWithAlgorithm(bimg, 8, cv2.CV_32S, cv2.CCL_SPAGHETTI)
# Make a mask with the edge label (0 is background, 1 is the first encountered label, i.e. the border)
ccedge = np.uint8((label_image != 1)*255)
cv2.imwrite("mask.png", ccedge, [cv2.IMWRITE_PNG_BILEVEL, 1])
# Zero every pixel touching the border
bimg = cv2.bitwise_and(bimg, ccedge)
# Remove border and invert again
bimg = 255 - bimg[1:h+1, 1:w+1]
# Save result
cv2.imwrite("result.png", bimg, [cv2.IMWRITE_PNG_BILEVEL, 1])

How to recolor image based on edge detection(canny)

I have a script, which is using for recoloring room walls based on color similarity. But I need to recolor a wall based on edge detection.
import cv2
import numpy as np
import sys
from PIL import Image
import numpy as np
from hex_to_rgb import color
def recolor(file_path, celor, lower_color, upper_color):
img = cv2.imread(file_path)
res = img.copy()
rgb = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
r2, g2, b2 = color(celor)
mask = cv2.inRange(rgb, lower_color, upper_color)
mask = mask/255
mask = mask.astype(np.bool)
res[:,:,:3][mask] = [b2, g2, r2] # opencv uses BGR
im_rgb = cv2.cvtColor(res, cv2.COLOR_BGR2RGB)
return im_rgb
file_path --> image
celor --> color, which you want to recolor
lower_color --> lower values of RGB
upper_color --> upper values of RGB

I am using Sobel edge detection to solve this problem. I tried with Canny edge detection also but it didn't give good results.
After edge detection, I applied the threshold to the image and found contours in the image. The problem here is that I am coloring the contour with the maximum area in this case. You will have to figure out a way to choose the contour you want to color.
img = cv2.imread("colourWall.jpg")
cImg = img.copy()
img = cv2.blur(img, (5, 5))
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
scale = 1
delta = 0
ddepth = cv.CV_16S
grad_x = cv.Sobel(gray, ddepth, 1, 0, ksize=3, scale=scale, delta=delta, borderType=cv.BORDER_DEFAULT)
grad_y = cv.Sobel(gray, ddepth, 0, 1, ksize=3, scale=scale, delta=delta, borderType=cv.BORDER_DEFAULT)
abs_grad_x = cv.convertScaleAbs(grad_x)
abs_grad_y = cv.convertScaleAbs(grad_y)
grad = cv.addWeighted(abs_grad_x, 0.5, abs_grad_y, 0.5, 0)
ret, thresh = cv2.threshold(grad, 10, 255, cv2.THRESH_BINARY_INV)
c, h = cv2.findContours(thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)
areas = [cv2.contourArea(c1) for c1 in c]
maxAreaIndex = areas.index(max(areas))
cv2.drawContours(cImg, c, maxAreaIndex, (255, 0, 0), -1)
plt.imshow(cImg)
plt.show()
Result:

How to optimize circle detection with Python OpenCV?

I have looked at several pages regarding optimizing circle detection using opencv in python. All seem to be specific to the individual circumstances of a given picture. What are some starting points for each of the parameters for cv2.HoughCircles? Since I am not sure what recommended values are, I have attempted looping over ranges but this is not producing any promising results. Why can't I detect any of the circles in this image?
import cv2
import numpy as np
image = cv2.imread('IMG_stack.png')
output = image.copy()
height, width = image.shape[:2]
maxWidth = int(width/10)
minWidth = int(width/20)
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
circles = cv2.HoughCircles(gray, cv2.HOUGH_GRADIENT, 1.2, 20,param1=50,param2=50,minRadius=minWidth,maxRadius=maxWidth)
if circles is not None:
# convert the (x, y) coordinates and radius of the circles to integers
circlesRound = np.round(circles[0, :]).astype("int")
# loop over the (x, y) coordinates and radius of the circles
for (x, y, r) in circlesRound:
cv2.circle(output, (x, y), r, (0, 255, 0), 4)
cv2.imwrite(filename = 'test.circleDraw.png', img = output)
cv2.imwrite(filename = 'test.circleDrawGray.png', img = gray)
else:
print ('No circles found')
This should be a straight forward circle detection, but all of the circles detected are not even close.

The main parameters that you should pay attention are minDist, minRadius and maxRadius.
Analyzing the radius first: you have an image that is 12 circles wide and 8 circles tall, which gives you a diameter of roughly width/12 for each circle, or a radius of (width/12)/2. The constraints that you have used allowed the algorithm to detect circles way bigger or smaller than necessary, therefore you should use a parameterization that is better fit for your image. In this case, I have used an interval [0.9 * radius, 1.1 * radius].
As there is no overlapping, you could say that the distance between two circles is at least the diameter, so minDist could be set to something like 2*minRadius.
This implementation is basically the same as yours, just updating those 3 parameters:
%matplotlib inline
import cv2
import numpy as np
import matplotlib.pyplot as plt
image = cv2.imread('data/balls.jpg')
output = image.copy()
height, width = image.shape[:2]
maxRadius = int(1.1*(width/12)/2)
minRadius = int(0.9*(width/12)/2)
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
circles = cv2.HoughCircles(image=gray,
method=cv2.HOUGH_GRADIENT,
dp=1.2,
minDist=2*minRadius,
param1=50,
param2=50,
minRadius=minRadius,
maxRadius=maxRadius
)
if circles is not None:
# convert the (x, y) coordinates and radius of the circles to integers
circlesRound = np.round(circles[0, :]).astype("int")
# loop over the (x, y) coordinates and radius of the circles
for (x, y, r) in circlesRound:
cv2.circle(output, (x, y), r, (0, 255, 0), 4)
plt.imshow(output)
else:
print ('No circles found')
The result is:

Normally circle detection can be done using traditional image processing methods such as thresholding + contour detection, hough circles, or contour fitting but since your circles are overlapping/touching, watershed segmentation may be better. Here's a good resource.
import cv2
import numpy as np
from skimage.feature import peak_local_max
from skimage.morphology import watershed
from scipy import ndimage
# Load in image, convert to gray scale, and Otsu's threshold
image = cv2.imread('1.jpg')
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
thresh = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)[1]
# Remove small noise by filtering using contour area
cnts = cv2.findContours(thresh, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
cnts = cnts[0] if len(cnts) == 2 else cnts[1]
for c in cnts:
if cv2.contourArea(c) < 1000:
cv2.drawContours(thresh,[c], 0, (0,0,0), -1)
cv2.imshow('thresh', thresh)
# Compute Euclidean distance from every binary pixel
# to the nearest zero pixel then find peaks
distance_map = ndimage.distance_transform_edt(thresh)
local_max = peak_local_max(distance_map, indices=False, min_distance=20, labels=thresh)
# Perform connected component analysis then apply Watershed
markers = ndimage.label(local_max, structure=np.ones((3, 3)))[0]
labels = watershed(-distance_map, markers, mask=thresh)
# Iterate through unique labels
for label in np.unique(labels):
if label == 0:
continue
# Create a mask
mask = np.zeros(gray.shape, dtype="uint8")
mask[labels == label] = 255
# Find contours and determine contour area
cnts = cv2.findContours(mask.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
cnts = cnts[0] if len(cnts) == 2 else cnts[1]
c = max(cnts, key=cv2.contourArea)
cv2.drawContours(image, [c], -1, (36,255,12), -1)
cv2.imshow('image', image)
cv2.waitKey()