There is no doubt about the fact that sometimes you have to work with documents containing graphic images with some text in a foreign language, which is subject to translation. About how to translate text from a picture or convert it into a readable format, now we will talk. It should be noted that this procedure includes several main stages, which need special attention.
How to translate text from a picture: options for action
To begin with, any program should initially recognize the text in a graphic format, after which determine the language in which it is written, and, finally, translate it. The main problem that most users face is not even recognition of text or language, but that most translation programs are not effective enough in terms of translation. In fact, this is the so-called machine translation, which is not welcomed by many, as it may contain errors related to the morphology of the language itself, used initially.
Nevertheless, if you try, for example, to translate the English text from the picture, the procedure may be as follows:
- converting graphics to text format;
- language recognition (optional);
- editing the original text (desirable, but not necessary);
- transfer.
Recognizing text with ABBYY FineReader
At the first stage, when the user has an image with text in a foreign language, it should be converted to plain text format. The software package ABBYY FineReader, which is considered a leader in its field, is best suited for this.
Recognition of the text is quite simple. In the program itself, you need to open the desired image file (or just drag it into the application field), then you need to press the recognition button and wait for the document scanning process to complete . If desired, you can set the language of the source document (if there are several, it is possible to set such a parameter as, for example, Russian + English).
Export
Next, you can export a fragment or the whole text. You can do this in the program itself very simply.
When the text is recognized, it can be copied for further pasting into any text editor or translator, but it is possible to quickly save it to any other format. For ease of use, it is better to choose the DOCX format, which is the main one for MS Word.
Insert Snippet into Text Editors
Now the text needs to be inserted into the editor. In the case of copying the selected text or fragment in FineReader to the Word editor, the paste is made from the clipboard with the corresponding command from the menu or with the key combination Ctrl + V (everyone knows that). If the initial text was saved as a file, you just need to open it using the menu command or the usual double-click on the document.
Now the main problem is how to translate text from Word from a picture? Alas, no way. This text editor is simply not intended for translations. But there is a way. The program allows you to install in your environment special add-ons ("Ruta", "Play", etc.). After installation, an additional section appears in the editor on the main panel and a button for quick access to the translation operation. It is enough to select the desired fragment and activate the translation.
How to translate text from a picture in a translator?
A fairly good translation method is the use of specialized programs or online services. One of the most powerful is the translate.ru service or a similar PROMPT translator program, which is installed on a computer. In any case, you need to insert the recognized text or fragment into the translator, indicate the direction of the translation and press the start button. Depending on the volume, translation may take some time. True, the online translator has restrictions on the number of characters that can be inserted into the main field for the source text. Moreover, it is a machine translation system. But in most cases, the translation is carried out not by individual words, but by whole phrases or sentences, taking into account the specifics of their construction and even idiomatic expressions.
You can, of course, turn to resources where "live" translators are involved in such operations, but, as a rule, all of them are paid, and translation time, based on the volume and complexity of the text, can take quite a lot. But in the programs, you can make, so to speak, a rough translation, and provided that you know some language, you can finish the editing yourself.
Using Screen Translator
But here is a program that translates source material from a picture into text, and from one language to another quickly and confidently without any transformations that were described above. The fact is that it already has built-in text recognition engines (Tesseract) and translation (Google Translate). How to translate text from an image using it? Very simple! To capture a text fragment, use the key combination Ctrl + Alt + Z, then, while holding down the left mouse button, the fragment of interest to the user is selected, and after a while the results appear - a window with recognized text and a window with a translation below.
The nuances of recognition and translation
How to translate text from a picture is already a little clear. Now a few words about additional actions at all stages. To get a high-quality translation, it is advisable to immediately edit the recognized text fragment (correct grammatically errors, remove extra spaces or format). In the future, this will simplify the work of the translator application, since some programs may perceive the same spaces as the end of a sentence.
It is recommended that machine translation be used solely as a draft, since texts of a specific orientation (technology, pharmaceuticals, etc.) cannot always be translated correctly due to the presence of many terms that are simply not available in the database of an online resource or program. But in the same service translate.ru or a similar stationary program, you can initially specify the area of use of the source text (computers, technology, medicine, etc.). This will simplify the task of the translator even more.
conclusions
That's all, in fact, with regard to the translation of texts from pictures. What to use? It seems that it is best to first make recognition, then edit, and then translate in the program. Screen Translator shows good results, but by and large, none of the currently created programs can translate text from one language to another fully correctly. This is due only to the fact that any language has its own morphological features.