here’s my code:

Is it possible to use OpenCV CUDA video writer and reader to write/read from a buffer allocated on the GPU memory instead of a file on the disk ?

Hi I’m quite new in OpenCV, I am currently trying to detect horizontal lines in an image using HoughLinesP function in opencv, in order to calculate the ratio between x and y.
The line detection result is shown in the second picture.

I have the following problem: I want to list around a hundred thousand products from AliExpress on my platform. Currently, I have reached 1000 products and I notice that I have accidentally listed some products twice. The reason is that different sellers on AliExpress offer the same product, each with their own titles. Therefore, it is not possible for me (also based on the title or other characteristics) to recognize whether I have already listed the item. The more products I list, the harder it becomes to keep track. I believe that once I reach ten thousand products, it will be nearly impossible to maintain an overview of what has already been listed.

Is there a cv2 function (in python) where I can update the display as cv2.waitKey(1) does, but without capturing keys?

I used a cam with the raspberryPI v. 4b with openvcv, but, when I start the program, tha stream start to lag and I can’t use it properly. I tried to use a sleep but nothing change. The code is this:
import cv2
import numpy as np
import time

I am trying to use the opencv_createsamples command but I keep getting the error command not found. I read somewhere that if I install an older version of opencv it will work. I tried many ways to do that. I might doing it wrong, but no matter what I try the opencv_createsamples command it not working.

Hello I’m using the following script where I (in the beginning) crop and resize the first 10 frames and write a new video with the result :

We have a QT based C++ application running on windows.
We are reading a data file which provides us images which we save in QPixmap variable.
When the data file is read in for loop huge number of image some 900 is received which we display in QLabel.
We want to save the images in .avi video for which we are using OpenCV API.

I am attempting to count the number of cars in traffic cameras. This is an obvious problem, and it’s been well-solved, assuming sufficient quality and resolution imagery. But for this use case I have 320×200 cameras, so … there just aren’t enough pixels for the standard algorithms to work well.