Ready to run the code right now on your Windows, macOS, or Linux systems?.Wanting to skip the hassle of fighting with the command line, package managers, and virtual environments?.Learning on your employer’s administratively locked system?.To learn how to resize an image using OpenCV and the cv2.resize method, just keep reading.įigure 3: Having trouble configuring your development environment? Want access to pre-configured Jupyter Notebooks running on Google Colab? Be sure to join PyImageSearch Plus - you will be up and running with this tutorial in a matter of minutes. Large resolution images are almost always downsampled to help image processing systems run faster and be more accurateīy the end of this tutorial, you’ll understand how to resize images with OpenCV using cv2 resize.High-resolution images are highly detailed - but from a computer vision/image processing perspective, we’re more interested in the structural components of the images, not so much the super fine-grained details.By definition, the larger the image, the more the data, and therefore the longer it takes for algorithms to process the data.Keep in mind that while high-resolution images are visually appealing to the human eyes, they harm computer vision and image processing pipelines: By decreasing the image size, we have fewer pixels to process (not to mention less “noise” to deal with), which leads to faster and more accurate image processing algorithms. It now looks “pixelated” and “blown up.”Īs I mentioned above, you’ll generally be decreasing the size of an image rather than increasing (exceptions do apply, of course). In the middle, we have resized the image to half its size - and other than the image being resized, there is no loss in image “quality.” However, on the right, we have dramatically increased the image size.

In terms of “quality” of the output image, decreasing an image’s size is always easier (and more aesthetically pleasing) than increasing it.

On the other hand, if we were to increase the size of the image, the interpolation function would have to “fill in the gaps” between pixels that previously did not exist.įor example, take a look at the image in Figure 2:įigure 2: Increasing and decreasing the size of an image with OpenCV. This is because the interpolation function simply has to remove pixels from an image. In general, it’s far more beneficial (and visually appealing) to decrease the size of the image. And the goal of an interpolation function is to take these neighborhoods of pixels and use them to either increase or decrease the size of the image. In this case, the “known points” are the pixels of our original image. The formal definition of interpolation is:Ī method of constructing new data points within the range of a discrete set of known data points. We also need to keep in mind the interpolation method of cv2.resize function. As we explore machine learning/deep learning techniques, we’ll find that our internal algorithms often ignore the aspect ratio of an image but more on that once we understand the fundamentals of computer vision. Exceptions most certainly do apply, though. In general, you’ll want to preserve the images’ aspect ratio when using cv2 resize - especially if these images are presented as output to the user.

They have been resized by ignoring the ratio of the width to the height of the image. And on the right, we have two images that have been distorted by not preserving the aspect ratio.

Ignoring the aspect ratio can lead to resized images that look compressed and distorted:įigure 1: Ignoring the aspect ratio of an image can lead to resized images that look distorted, squished, and crunched. When resizing an image, it’s important to keep in mind the aspect ratio - which is the ratio of an image’s width to its height. You can start by choosing your own datasets or using our PyimageSearch’s assorted library of useful datasets.īring data in any of 40+ formats to Roboflow, train using any state-of-the-art model architectures, deploy across multiple platforms (API, NVIDIA, browser, iOS, etc), and connect to applications or 3rd party tools. Sign up or Log in to your Roboflow account to access state of the art dataset libaries and revolutionize your computer vision pipeline. Roboflow has free tools for each stage of the computer vision pipeline that will streamline your workflows and supercharge your productivity. This prepares us to handle images of different sizes in real-world applications. Scaling, or simply resizing, is the process of increasing or decreasing the size of an image in terms of width and height.Ī dataset with diverse image sizes and dimensions is beneficial in understanding the function and effects of the cv2.resize function in OpenCV.

Change imagesize