CZUR ET24 Pro Professional Book Scanner, 24MP Document Camera, 3rd Gen Auto-Flatten & Deskew Tech, A3 Document Scanner, 180+ Languages OCR, Support HDMI, for Windows/MacOS/Linux

I had previously tried several other book scanners that look a lot like this one, but they had so many issues and problems that they were completely useless. This one, however, works right out of the box. But, if I actually knew of another similar scanner that worked to compare this against, I would probably only rate this 4 stars instead of 5 (keep reading to see why).First off, a big "Thank You" for supporting Linux. Even though these book scanners are nothing more than a camera and there's plenty of open source software on Linux to do all of the necessary functions for book scanning, it's still nice to be able to install the official software for a device and just use it.The line-lasers on this scanner help a lot with page flattening and is one major thing that similar scanners are missing. Sure page flattening can be done entirely in software, but having a line-laser really improves the algorithm and makes for better results.Unlike other similar scanners, this one seems to correctly report itself in the USB enumerations for all of the resolutions and settings. That is a must for compatibility across platforms. And, since it appears as a standard video and camera source device, it means you can easily use this with 3rd party software if desired.The finger cots are a little interesting. They do work to let you hold down the book and have special markings so that the software can remove them. However, the software doesn't fully remove the cots from the captured picture. It mostly uses them to figure out the boundary of the book so it can remove your hands/arms from the picture. There is still an outline of the cots on the page and it will generally cause some stray character interpretations and formatting weirdness in the OCR step -- and seemingly worse on the right-hand side of the page than the left-hand. For example, it will usually see the curvature of the cot as a ")" character and its presence will often cause a paragraph break or formatting change to the text it's recognizing that it wouldn't if the cot wasn't in the picture. I think they could have done a little more post-processing to remove them better, but they do generally work.The 24MP camera is more than sufficient for adequately scanning pages for OCR and to capture decent images from pictures in books. However, I wouldn't go with any lower resolution. OCR is very dependent on the pixel count to font size ratio (essentially the DPI captured). It has to be somewhat close to what the OCR engine was trained with or you will get garbage out. I learned that one years ago when trying to do an OCR of a book captured to fairly low-res JPG files. All I got out of OCR was random noise and so I figured it had no chance due to the JPGs being too low-res. Then a coworker clued me in on the pixel count to size ratio requirements and I used imagemagick to resample all of the files to a higher pixel count (still same basic "camera" resolution and same source images only more pixels) and tried the OCR again. This time it yielded a nearly 100% perfect OCR output and totally amazed me.I mention that about OCR and pixels because the resolution of the camera for a scanner like this is important. If your camera is too low-res, you may find you have to first resample all images before you can OCR it. I'm finding that the 24MP on this scanner is sufficiently adequate for the majority of fonts in books, but just barely. So, you definitely don't want to go lower than that.My first big complaint on this scanner is the user interface of the software. Yes, it works and works on Linux and is written in Qt making it a very compatible and cross-platform tool (albeit Qt 4, which is quite ancient since Qt 6 is the current standard), but it's very clunky, not overly intuitive, and somewhat annoying. It doesn't remember the last saved path correctly when relaunching. It has no easy way to automatically operate on multiple subfolders (such as if you want to scan different chapters to different folders), and there's a number of bugs and limitations in the software. However, it was usable to scan the books I needed to scan without resorting to pulling in 3rd party tools.One note on the software -- if you run an OCR session, the tool launches off a separate helper thread that then calls the OCR tool. That in itself would be fine, except that when its done with the extra thread, it doesn't close down that thread. When you are done and exit the application completely, that extra thread causes the app to not fully shutdown and to just hang. And it hangs with the scanner device locked in-use, meaning that if you relaunch the scanner software it can't find the scanner camera device anymore since it's "already in use". To get around this, you have to find the process ID for the stuck app and manually kill it after you close the app (or reboot). It only hangs, though, if you do an OCR session, and works fine if just using it to capture scanned pages.My second complaint is that the lights, even with the extra light-bar attached, aren't fully sufficient to evenly illuminate the page and causes some shadows. If you are scanning to OCR, it's fine, but if you were doing archival quality scans, it's not -- you will need to add some extra lighting. Perhaps this is where their extra light/shadow box is useful?The little LCD screen on top is basically useless, or at least it was for me. It's not moveable and is very small for trying to see a preview image. If you are sitting down at the table in front of the scanner as I was, since it isn't moveable, you can't see it at all. So I really have no idea what it was showing as I couldn't see it. For previews, I just used the live video feed on the computer in the scanner software to make sure things were lined up. They should have either saved themselves a few dollars per unit and just not put a display on it or spent a few dollars more per unit and put a larger, moveable screen on it that would actually be useful, but this one is useless.I mostly use the footswitch to trigger page captures. However, you need to keep an eye on either the red light at the top of the scanner or the images captured in the software to make sure it took a snapshot. A number of times, I would press the footswitch and then realized that nothing happened. I don't know if it's a problem with the footswitch just not sending a signal or if it's the debouncing logic on the switch or something else. But occasionally it wouldn't trigger. However, once you get in your rhythm of flipping pages, double checking the alignment periodically, and clicking the footswitch, you can scan a large book very quickly -- much faster than any flatbed scanner by a long shot. Though don't be concerned if the tiny preview images of each captured page occasionally look a little wonky -- that's just a Qt software quirk on how it renders such preview images (as I've seen it in other Qt-based apps). And it still looks good enough to be able to get a general sense of the page scan quality.I haven't tried the HDMI output connector nor have I used the microphone and "presentation" mode. At first, it seemed a bit of a gimmick, especially the microphone, but then I realized you could use it with things like OBS and either do a live stream presentation or capture a video presentation, and that could be useful.I primarily got the scanner because I have a number of books in foreign languages that I need to scan and OCR so that I can translate them to be able to read them. Several of the books are rare and need to be handled with care. I didn't want to scan them in a flatbed scanner, as doing so usually destroys the spine of the book. This device fits the bill and has worked quite well.For OCR, it comes with Abbyy, which is a commercial grade OCR tool. I had previously used the open-source Tesseract quite a bit and found it to be absolutely fantastic and ran circles around commercial OCR apps I had tried in the past. So, it was interesting to see how Abbyy compared. I must say, I'm pretty impressed. In fact, I would be hard pressed to tell the two apart. I personally think that Abbyy just took the open-source Tesseract OCR engine and trained it with their own datasets to make it into a separate commercial app. They seem to be too similar to think otherwise. But that also means you will have no issues with OCR. The Abbyy OCR this comes with works fantastically. However, if you ever prefer to use something different, you can always take the JPG files from this scanner and run them through Tesseract or other OCR tool.You can also inject your own scripts into the OCR batch process. That means if you want to run things through ScanTailor or something first, you certainly can.That's pretty much it -- this is not a perfect scanner, but it comes with everything you need to get going, was very nicely packaged, and worked right out of the box and on my Linux computers. I had previously tried to save a few bucks and tried several competitor scanners, but couldn't ever get them to work correctly. If there was anything else to actually compare this with, I would give it 4 stars instead of 5 because of the minor quirks and bugs -- their software isn't fully polished. But for general scanning needs, you won't go wrong with this scanner.

I've been using the CZ24 scanner for a few months now and have been pleased with the output. I couldn't decide between this and the ScanSnap SV600 - and ended up buying both. There are strengths and weaknesses to both, but I find myself using the CZ24 far more often of the two. For most book scanning jobs, this does everything I need. The quality of high color saturation scans is better with the SV600, but that is only a fraction of my typical use. For the color you generally encounter in most books, publications, journals, etc., the CZ24 is more than adequate.If I had to do it over, I would have bought the CZ24 alone. The CZ24 is the more versatile of the two and the CZUR software is easier to use. Easier - but not highly intuitive. Spend some time and play with the software to understand how it works before dedicating yourself to scanning a 200 page book. Once understood, it goes pretty well.

The CZUR ET24 Pro is a game-changer for digitizing books and documents! Its 24MP camera captures crisp, high-quality scans, and the 3rd Gen Auto-Flatten & Deskew technology works like magic, ensuring even curved pages come out perfectly flat. I love the A3-sized scanning capability—great for oversized documents. The HDMI support is a nice touch for presentations, and compatibility with Windows, MacOS, and Linux makes it versatile.The OCR software, while functional with 180+ languages, feels a bit dated. However, pairing the scans with modern LLMs easily bridges this gap, delivering top-notch text recognition. Setup was a breeze, and the scanner’s sleek design fits nicely on my desk. Highly recommend for professionals, students, or anyone needing reliable, high-quality scanning!

Short answer is it works very well with books that have straight flat pages i.e. the pages that open symmetrically with same curvature left to right so the laser beams have symmetrical tracing lines on both pages. The readability of text and quality of black & white text is really great The quick scanning promised by this scanner works only if your priority is raw digitalization of content. If you're hoping for even margins, straight pages then that requires much more love & care and perhaps not with czur software but photoshop.The problems I experienced:1. Thick books like dictionaries. If you scan the middle of the book with pages equally spread to left and right the scan is great. The challenge is with the first and last pages [that is likely 80% of the book content] when the book would not stay open or symmetrically open and the pages on left and right do not have same curvature which is essential for the flattening laser beams operation.. In result even if the left and right pages come out flat they come out with different margins. I will try the czur assistive cover to see if it helps.2. There is a problem to get a nice scan with used books [especially the smaller format pages] with pages stubborn to stay flat. Not flat as left to right but top to bottom. If the page is wavy across all 3D axis the scan will not come out well. It would be nice to have some feature to hold the book page down around the edges to symmetrize the left & right pages or perhaps a transparent press down page. Those 2 yellow finger holders are not effective in that case, perhaps a holder that would pin down the page along the page edge would work better.3. The CZUR software has the auxiliary correction feature that as I understand offers the option to manually adjust the pages curvature lines to improve the flattening effect (see problem 1 &2 above). I spent some time with it and cannot figure out how it works toward the expected improved result, nor I found any good tutorial for that feature on czur pages or in any reviews. If it works indeed this is a last resort option anyway as this operation must be applied to each individual page in the book, let's hope czur will provide more documentation for that.4. The color pages scans seem to suffer from glare. The scan is not always truthful to the original. I tried to play with the adjustments such as color, brightness and contrast Perhaps it requires even more finesse on my end or better ambience lighting like the one used in professional photography.