Recently, I received a small project that involved organizing the vector boundaries of rivers. The task is quite simple as the data is available; however, the key issue is that the provided data was collected around 2010 and is only available in hard copy, with no electronic version. The coordinates are organized in columns, covering dozens of images with nearly 2000 coordinates. Manually typing them would be time-consuming and labor-intensive, so the best approach is to find a suitable optical character recognition (OCR) tool to recognize the text and then organize it. After checking online recommendations, I manually downloaded and installed nearly ten different tools. To be honest, I didn’t find any particularly good software. While these tools perform reasonably well for standard text recognition, they fall short for columnar tabular images.
-
WeChat and QQ Built-in Text Recognition
The operation is very simple. Just send the captured image to someone, then click on the image, select the dot in the lower right corner, and choose 【Extract Text】. The effect is acceptable; normal printed text can be recognized, while handwriting recognition depends on how well it is written, and it cannot perform batch recognition. Recommended
-
Built-in Scanner on Phones
Many smartphones now come with built-in scanners that can recognize text. The operation is similar to WeChat, but it saves a step of sending the image. Just take a photo and scan it immediately, which is quick and suitable for small scenarios requiring fast text recognition. Recommended
-
Tianruo OCR
This software is paid, and while its interface looks simple, it requires an internet connection and has very few recognition functions. Not Recommended
-
Shangshu No. 7 OCR
The operation is quite unconventional, and while the text recognition is acceptable, it does not handle tables well. Not Recommended
-
Lightning OCR
This one also requires a membership to use, and I haven’t tried its effectiveness. Not Recommended
-
ABBYY FineReader
To speak plainly, this software provides the best text recognition and column mode recognition among the few I have tried. This software is relatively large, over 500MB, and the only downside is that it is not free; the trial allows a maximum of 9 photos. I saw that it wasn’t free, tried it, and then uninstalled it without taking a screenshot. However, for high-end users who are not short on cash, I highly recommend this software. Recommended
-
CAJViewer
This is a reader that can recognize PDFs and other formats but cannot recognize images. Of course, if you convert the image to PDF, it can recognize it, which is somewhat cumbersome. It is suitable for situations where you want to copy a portion of text while reading a PDF. Generally Recommended
-
Zhuan Yi Xia
As per my usual practice, the most recommended software always comes last. This software is free, has many functions, and performs well in text recognition. Although it still does not fully meet my needs for recognizing column coordinates, it is indeed very suitable for everyday image text recognition. The main interface is as follows:
Although the upper right corner has the word “Login”, not logging in does not affect the functionality. It can recognize images, directly recognize web pages or desktop screenshots, and can output to Word and Excel.
It can also recognize various types of identification documents, as shown in the following image:
It accurately and quickly recognizes ID card numbers, bank card numbers, etc. If you need to recognize a lot of identity information, this software is absolutely my top recommendation. Highly Recommended, reply with 【Zhuan Yi Xia】 to get it directly
There are also a few other options, such as Fast OCR, OneNote, etc., as well as various online recognition tools like Baidu and Tencent. However, online tools have a major drawback of being significantly affected by network conditions and are not free. Therefore, using an offline tool is the best choice.