Extract dark contour - opencv

I want to extract the darker contours from images with opencv. I have tried using a simple threshold such as below (c++)
cv::threshold(gray, output, threshold, 255, THRESH_BINARY_INV);
I can iterate threshold lets say from 50 ~ 200
then I can get the darker contours in the middle
for images with a clear distinction such as this
here is the result of the threshold
but if the contours near the border, the threshold will fail because the pixel almost the same.
for example like this image.
What i want to ask is there any technique in opencv that can extract darker contour in the middle of image even though the contour reach the border and having almost the same pixel as the border?
(updated)
after threshold darker contour in the middle overlapped with border top.
It makes me fail to extract character such as the first two "SS".

I think you can simply add a edge preserving smoothing step to solve this :
// read input image
Mat inputImg = imread("test2.tif", IMREAD_GRAYSCALE);
Mat filteredImg;
bilateralFilter(inputImg, filteredImg, 5, 60, 20);
// compute laplacian
Mat laplaceImg;
Laplacian(filteredImg, laplaceImg, CV_16S, 1);
// threshold
Mat resImg;
threshold(laplaceImg, resImg, 10, 1, THRESH_BINARY);
// write result
imwrite("res2.tif", resImg);
This will give you the following result : result
Regards,

I think using laplacian could partialy solve your problem :
// read input image
Mat inputImg = imread("test2.tif", IMREAD_GRAYSCALE);
// compute laplacian
Mat laplaceImg;
Laplacian(inputImg, laplaceImg, CV_16S, 1);
Mat resImg;
threshold(laplaceImg, resImg, 30, 1, THRESH_BINARY);
// write result
imwrite("res2.tif", resImg);
Using this code you should obtain something like :
this result
You can then play with final threshold value and with laplacian kernel size.
You will probably have to remove small artifacts after this operation.
Regards

Related

Remove Boxes/rectangles from image

I have the following image.
this image
I would like to remove the orange boxes/rectangle around numbers and keep the original image clean without any orange grid/rectangle.
Below is my current code but it does not remove it.
Mat mask = new Mat();
Mat src = new Mat();
src = Imgcodecs.imread("enveloppe.jpg",Imgcodecs.CV_LOAD_IMAGE_COLOR);
Imgproc.cvtColor(src, hsvMat, Imgproc.COLOR_BGR2HSV);
Scalar lowerThreshold = new Scalar(0, 50, 50);
Scalar upperThreshold = new Scalar(25, 255, 255);
Mat mask = new Mat();
Core.inRange(hsvMat, lowerThreshold, upperThreshold, mask);
//src.setTo(new scalar(255,255,255),mask);
what to do next ?
How can i remove the orange boxes/rectangle from the original images ?
Update:
For information , the mask contains exactly all the boxes/rectangle that i want to remove. I don't know how to use this mask to remove boxes/rectangle from the source (src) image as if they were not present.
This is what I did to solve the problem. I solved the problem in C++ and I used OpenCV.
Part 1: Find box candidates
Firstly I wanted to isolate the signal that was specific for red channel. I splitted the image into three channels. I then subtracted the red channel from blue channel and the red from green channel. After that I subtracted both previous subtraction results from one another. The final subtraction result is shown on the image below.
using namespace cv;
using namespace std;
Mat src_rgb = imread("image.jpg");
std::vector<Mat> channels;
split(src_rgb, channels);
Mat diff_rb, diff_rg;
subtract(channels[2], channels[0], diff_rb);
subtract(channels[2], channels[1], diff_rg);
Mat diff;
subtract(diff_rb, diff_rg, diff);
My next goal was to divide the parts of obtained image into separate "groups". To do that, I smoothed the image a little bit with a Gaussian filter. Then I applied a threshold to obtain a binary image; finally I looked for external contours within that image.
GaussianBlur(diff, diff, cv::Size(11, 11), 2.0, 2.0);
threshold(diff, diff, 5, 255, THRESH_BINARY);
vector<vector<Point>> contours;
findContours(diff, contours, CV_RETR_EXTERNAL, CV_CHAIN_APPROX_NONE);
Click to see subtraction result, Gaussian blurred image, thresholded image and detected contours.
Part 2: Inspect box candidates
After that, I had to make an estimate whether the interior of each contour contained a number or something else. I made an assumption that numbers will always be printed with black ink and that they will have sharp edges. Therefore I took a blue channel image and I applied just a little bit of Gaussian smoothing and convolved it with a Laplacian operator.
Mat blurred_ch2;
GaussianBlur(channels[2], blurred_ch2, cv::Size(7, 7), 1, 1);
Mat laplace_result;
Laplacian(blurred_ch2, laplace_result, -1, 1);
I then took the resulting image and applied the following procedure for every contour separately. I computed a standard deviation of the pixel values within the contour interior. Standard deviation was high inside the contours that surrounded numbers; and it was low inside the two contours that surrounded the dog's head and the letters on top of the stamp.
That is why I could appliy the standard deviation threshold. Standard deviation was approx. twice larger for contours containing numbers so this was an easy way to only select the contours that contained numbers. Then I drew the contour interior mask. I used erosion and subtraction to obtain the "box edge mask".
The final step was fairly easy. I computed an estimate of average pixel value nearby the box on every channel of the image. Then I changed all pixel values under the "box edge mask" to those values on every channel. After I repeated that procedure for every box contour, I merged all three channels into one.
Mat mask(src_rgb.size(), CV_8UC1);
for (int i = 0; i < contours.size(); ++i)
{
mask.setTo(0);
drawContours(mask, contours, i, cv::Scalar(200), -1);
Scalar mean, stdev;
meanStdDev(laplace_result, mean, stdev, mask);
if (stdev.val[0] < 10.0) continue;
Mat eroded;
erode(mask, eroded, cv::Mat(), cv::Point(-1, -1), 6);
subtract(mask, eroded, mask);
for (int c = 0; c < src_rgb.channels(); ++c)
{
erode(mask, eroded, cv::Mat());
subtract(mask, eroded, eroded);
Scalar mean, stdev;
meanStdDev(channels[c], mean, stdev, eroded);
channels[c].setTo(mean, mask);
}
}
Mat final_result;
merge(channels, final_result);
imshow("Final Result", final_result);
Click to see red channel of the image, the result of convolution with Laplacian operator, drawn mask of the box edges and the final result.
Please note
This code is far from being optimal, especially the last loop does quite a lot of unnecessary work. But I think that in this case readability is more important (and the author of the question did not request an optimized solution anyway).
Looking towards more general solution
After I posted the initial reply, the author of the question noted that the digits can be of any color and their edges are not necessarily sharp. That means that above procedure can fail because of various reasons. I altered the input image so that it contains different kinds of numbers (click to see the image) and you can run my algorithm on this input and analyze what goes wrong.
The way I see it, one of these approaches is needed (or perhaps a mixture of both) to obtain a more "general" solution:
concentrate only on rectangle shape and color (confirm that the box candidate is really an orange box and remove it regardless of what is inside)
concentrate on numbers only (run a proper number detection algorithm inside the interior of every box candidate; if it contains a single number, remove the box)
I will give a trivial example of the first approach. If you can assume that orange box size will always be the same, just check the box size instead of standard deviation of the signal in the last loop of the algorithm:
Rect rect = boundingRect(contours[i]);
float area = rect.area();
if (area < 1000 || area > 1200) continue;
Warning: actual area of rectangles is around 600Px^2, but I took into account the Gaussian Blurring, which caused the contour to expand. Please also note that if you use this approach you no longer need to perform blurring or laplace operations on blue channel image.
You can also add other simple constraints to that condition; ratio between width and height is the first one that comes to my mind. Geometric properties can also be a good option (right angles, straight edges, convexness ...).

OpenCV - Separating letter contours in text segmentation

I’ve been working with text recognition in a dataset of images. I want to segment the characters of the image using components and finding contours of a thresholded image. However, many of the characters are merged with each other and with other components in the image.
Can you give me some idea for separating them? Thanks for the help!
Below are some examples, and part of my code:
Mat placa_contornos = processContourns(img_placa_adaptativeTreshold_mean);
vector<vector<Point>> contours_placa;
findContours(placa_contornos,
contours_placa,
CV_RETR_EXTERNAL, externos)
CV_CHAIN_APPROX_NONE);
vector<vector<Point> >::iterator itc = contours_placa.begin();
while (itc != contours_placa.end()) {
//Create bounding rect of object
Rect mr = boundingRect(Mat(*itc));
rectangle(imagem_placa_cor, mr, Scalar(0, 255, 0));
++itc;
}
imshow("placa con rectangles", imagem_placa_cor);
Results examples
original image, binarized image, result
I would try to erode the binary image more to see if that helps. You may also want to try fixing the skew and then removing the bottom line that connects the letters
Also, this might be relevant: Recognize the characters of license plate
You can try an opening operation on your thresholded image to get rid of the noise. You can adjust the kernel size based on your need.
// Get a rectangular kernel with size 7
Mat element = getStructuringElement(0, Size(7, 7), Point(1, 1));
// Apply the morphology operation
morphologyEx(placa_contornos, result, CV_MORPH_OPEN, element);
It gives the following intermediate output on your thresholded image, I guess it would improve your detection.

How do I create a histogram which can be used for calcBackProject in opencv?

I have specified the histogram as
MatND skinCrCbHist =Mat::zeros(Size(256,256),CV_8UC1);
ellipse(skinCrCbHist, Point(113, 155.6), Size(283.4, 159.2), 43.0, 0.0, 360.0, Scalar(255), -1); // Using a really big ellipse to find any sort of back projection in CrCb domain.
cvtColor(src, ycrcb, CV_BGR2YCrCb); //src is input, image of a person
float crrange[]={0,255};
float cbrange[]={0,255};
const float* ranges[]={crrange,cbrange};
int channelsy[]={1,2};
calcBackProject( &ycrcb, 1, channelsy, skinCrCbHist, backproj, ranges, 255, true );
imshow("bp",backproj);
The problem i face is that backproj shows a completely black image.
When I used a normal histogram created with calcHist on a natural image, i do get some sort of backprojection. But how do i use a histogram, i create artificially, by specifying an ellipse, to get a backprojection.
If I understood your problem correctly, you could use mask with the original calcHist function.
You didn't specified which version of OpenCV you are using, so I will assume the latest 2.4.6.0. The method prototype is following (omitting defaults, and types):
calcHist(images, nimages, channels, mask, hist, dims, histSize, ranges)
The third parameter is mask. The mask means, that the function will ignore all pixels which matches zero pixels in mask. In program the mask is another image correctly setup.
Here is pseudo-code for you problem:
1) get input image
2) create matrix of same size as input of type CV_8UC1 filled with zeros
3) draw white (value 255) ellipse on the new image
4) call caclHist with the new image as mask
http://docs.opencv.org/modules/imgproc/doc/histograms.html

Image processing techniques to stand out white tape on the floor with opencv

I have the following image:
And I'd like to obtain a thresholded image where only the tape is white, and the whole background is black.. so far I've tried this:
Mat image = Highgui.imread("C:/bezier/0.JPG");
Mat byn = new Mat();
Imgproc.cvtColor(image, byn, Imgproc.COLOR_BGR2GRAY);
Mat thresh = new Mat();
// apply filters
Imgproc.blur(byn, byn, new Size(2, 2));
Imgproc.threshold(byn, thresh, 0, 255, Imgproc.THRESH_BINARY+Imgproc.THRESH_OTSU);
Imgproc.erode(thresh, thresh, Imgproc.getStructuringElement(Imgproc.MORPH_RECT, new Size(4, 4)));
But I obtain this image, that is far away from what I want:
The tape would be always of the same color (white) and width (about 2cm), any idea? Thanks
Let's see what you know:
The tape has a lower contrast
The tape is lighter than the background
If you know the scale of the picture, you can run adaptive thresholds on two levels. Let's say that the width of the tape is 100 pixels:
Reject a pixel that has brightness outside of +/- x from the average brightness in the 50x50 (maybe smaller, but not larger) window surrounding it AND
Reject a pixel that has brightness smaller than y + the average brightness in the 100x100(maybe larger, but not smaller) window surrounding it.
You should also experiment a bit, trying both mean and median as definitions of "average" for each threshold.
From there on you should have a much better-defined image, and you can remove all but the largest contour (presumably the trail)
I think you are not taking advantage of the fact that the tape is white (and the floor is in a shade of brown).
Rather than converting to grayscale with cvtColor(src, dst, Imgproc.COLOR_BGR2GRAY) try using a custom operation that penalizes saturation... Maybe something like converting to HSV and let G = V * (1-S).

Adaptive threshold of blurry image

I have a fairly blurry 432x432 image of a Sudoku puzzle that doesn't adaptively threshold well (take the mean over a block size of 5x5 pixels, then subtract 2):
As you can see, the digits are slightly distorted, there are a lot of breakages in them, and a few 5s have fused into 6s and 6s into 8s. Also, there's a ton of noise. To fix the noise, I have to make the image even blurrier using a Gaussian blur. However, even a fairly large Gaussian kernel and adaptive threshold blockSize (21x21, subtract 2) fails to remove all the breakages and fuses the digits together even more:
I've also tried dilating the image after thresholding, which has a similar effect to increasing the blockSize; and sharpening the image, which doesn't do much one way or the other. What else should I try?
A pretty good solution is to use morphological closing to make the brightness uniform and then use a regular (non-adaptive) Otsu threshold:
// Divide the image by its morphologically closed counterpart
Mat kernel = Imgproc.getStructuringElement(Imgproc.MORPH_ELLIPSE, new Size(19,19));
Mat closed = new Mat();
Imgproc.morphologyEx(image, closed, Imgproc.MORPH_CLOSE, kernel);
image.convertTo(image, CvType.CV_32F); // divide requires floating-point
Core.divide(image, closed, image, 1, CvType.CV_32F);
Core.normalize(image, image, 0, 255, Core.NORM_MINMAX);
image.convertTo(image, CvType.CV_8UC1); // convert back to unsigned int
// Threshold each block (3x3 grid) of the image separately to
// correct for minor differences in contrast across the image.
for (int i = 0; i < 3; i++) {
for (int j = 0; j < 3; j++) {
Mat block = image.rowRange(144*i, 144*(i+1)).colRange(144*j, 144*(j+1));
Imgproc.threshold(block, block, -1, 255, Imgproc.THRESH_BINARY_INV+Imgproc.THRESH_OTSU);
}
}
Result:
Take a look at Smoothing Images OpenCV tutorial. Except GaussianBlur there are also medianBlur and bilateralFilter which you can also use to reduce noise. I've got this image from your source image (top right):
Update: And the following image I got after removing small contours:
Update: also you can sharpen image (for example, using Laplacian). Look at this discussion.
Always apply gaussian for better results.
cvAdaptiveThreshold(original_image, thresh_image, 255,
CV_ADAPTIVE_THRESH_GAUSSIAN_C, CV_THRESH_BINARY, 11, 2);

Resources