No. You can see for yourself by inspecting your browser's outgoing network requests. With AutoCropper, you can upload a 20 megabyte scan and download its automatically separated and perfectly cropped images at 100% full quality in seconds. That's a pretty big image file. Uploading that image file to a server would take seriously fast upload speeds.
AutoCropper is fast because your full size scan is not uploaded, sent to any server, or saved anywhere. What is sent to the server is a significantly compressed, colorless, metadata-stripped copy of your scan. For a 20 megabyte scan, a 240 kilobyte binary copy is sent to the server (around 5% its original size). This is necessary in order to detect the individual images on your scan using a computer vision machine learning algorithm I created myself. You can read in detail exactly how AutoCropper works step-by-step, from the moment you upload a file to everything in between.
In fact, 100% of the top 10 results on Google for "crop image online" actually store your full size upload, because that is how they make money. So I'm actually better than those, and I'm explaining to you how it all works!
The biggest concern using an web application like this is knowing what's going on behind the scenes, and so I would like to show you that AutoCropper is a great and safe tool. Read on to see actual examples of what is sent for a given upload (using my actual scanned photos!).
When you upload your scan or multiple scans to AutoCropper, the original raw image is stored in your browser's local storage. Locally, entirely in your browser. This is done so when you click "Export Photos", the raw image can be cropped client-side. In this example we are using this image, which is a whopping 16mb and 3400 × 4680: example courtesy of yours truly.
A copy is made that is a radically compressed black and white version of the raw upload. In this case, the 16 mb raw scan is compressed into a 0.24 mb image file that is 363 by 500 pixels. Its dimensions much lower quality, black and white copy is sent for processing. The reason is because the separate images on your scan can be identified by my algorithm just as confidently at just 10% quality. You can see the image sent to the server here. At 3% quality and file size, you can still clearly identify the whitespace separating the images.
In response, the server returns a list of coordinates, one for each separate image found on the scan. Each coordinate is an X, Y position of the top left corner of the image, along with its height and width. These coordinates are obviously relative to the size of the tiny version sent to the server. So, when the server responds, we simply multiply these coordinates by a scaler, in this case the scaler is the small image's height (500) divided by the original image's height (4680), or 9.46.
Read more about how AutoCropper works in detail.
Storing the scans you upload would not make sense for me or you. It would be incredibly inefficient both in terms of time for you, and also cost for me.
Storing random people's full size scans that are uploaded here would not only be weird (which is not that convincing because there are plenty of weird people out there), but also cost me money because scans are large in file size. I also have no idea what I'd do with them. I am only putting so much emphasis here because privacy is a very legitimate concern for anyone using a website like this.
Here are the main reasons AutoCropper does not save your scans:
In terms of storage, I would hope your scans are large in size, as in 20+ megabytes each. If they aren't, I would suggest increasing your scanner's DPI and quality settings. Read more for scanning tips here.
With AutoCropper, you can upload a batch of 10 scans at once (or 25 if you purchase a lifetime license), which comes out to 200 megabytes.
Although storage costs continue to become cheaper, the cost of storing this data and the liability associated with doing so would grow very quickly. Liability alone with GDPR and such would be a massive headache.
As I have explained, AutoCropper is very fast. Your 10 scan 200 megabyte batch upload will be processed and ready for review in about 5 to 10 seconds. That is because each scan is compressed to 5% its original size, to a maximum of 250 kilobytes, so the total data sent to my API is just 2 megabytes. Read more about how AutoCropper achieves these fast results and how the compressed scan is sent to my server.
Storing and uploading your full sized scans behind the scenes would place a considerable load on an already complex web application like AutoCropper. In fact, the biggest delay between your clicking Upload and your crops being displayed, is your full size scans being stored client side. They need to be stored client side (in your browser - not in any cloud) in order for AutoCropper to export the full size crops when you click the Export button.
I truly do not know what I would do with them. It's not like they would be worth anything when you get datasets of millions of images, tagged and everything for free.
With AutoCropper, you can upload a batch of 10 scans at once (or 25 if you purchase a lifetime license!). Let's say each scan is around 20 megabytes. That would require some very fast upload speeds to secretly send 200 megabytes of image data to a server. You'd also be able to see this upload very clearly in your browser's developer tools. Learn more about how to see outgoing network requests to see for yourself.
In 2023, the average person's internet upload speed is 8.5 megabits per second, or about 1 megabyte per second. Uploading your scans to some secret server in this example would take about 3 minutes, and require you to stay on AutoCropper that entire time, which is very unlikely.
Luckily as I have explained, AutoCropper is very fast. Your 10 scan 200 megabyte batch upload will be processed and ready for review in about 5 to 10 seconds. Unless you stick around for an additional 5 minutes, I don't think there is a way for me to secretly upload your scans in the background. Either way, AutoCropper does not save or store your uploads anywhere. You can inspect the network tab in your browser to see the outgoing requests and their sizes.
Last but not least and this one is the most important - it would be really weird to store your scans, and also probably illegal. Even weirder especially after having written all of this content trying to prove you otherwise.
