Prologue
Today, while reinstalling the OS on my laptop, I decided to read smart books on my phone running Android OS.
For this purpose, I installed the well-known CoolReader. But bad luck, the book I wanted to read was not on my phone. What to do? Download the book you need from an online library, because the reader has a set of online libraries made specifically so that the user does not wander the Internet looking for a book and does not waste his time on it. Scrolled to the section with online libraries...
Are you one of those people who just watch a movie once, who reads a book once and then puts it on the shelf forever or in electronic format in the “read” folder or just delete it? Will you buy unnecessary, expensive, unfavorable books? Under what conditions e-books are borrowed, you can find out from the source, but in general the discount can be up to 50% of the purchase price.
Books are currently borrowed for 21 days, starting from the time your order is received. And there are also one-time packages where you can get up to 49 CZK per title. With powerful hardware, the browser is also good for speeds, and it doesn't really matter that it's not possible to target the output to the width of the display. Browsing multiple contrasting sites is a problem, just like with any other reading device e-books. Where there is White background and black font on the page, that's normal, of course.
The essence of the problem:
Free downloading and reading books directly from the “CoolReader” reader for Android.
At some point, I installed CoolReader and started searching in the menu items for the ability to search and download books for free. After all, it is quite logical that in a popular e-reader on Android, this function is literally vital.
It turned out that until a certain time the Flibusta library was available in the reading room. It was possible to go to this library, download and read a book immediately after installing the reader, but this did not last very long.
There is nothing important in the settings, except perhaps a nice option for setting the image on the display. Even two icons on the home screen can be replaced by another application. Just like with tablets or smartphones, you can also install your own keyboard and reinstall it. Without the worse backlight uniformity, this reader would be almost error-free. As “some China,” the reader is well prepared. It's just the size and weight it stores, uses and transfers. The hardware is top-notch, the reader works great, the environment is fast, the pages flicker.
After about 10 minutes of active button presses, I did it, added Flibusta to CoolReader.
And now point by point and with pictures:
1. Open CoolReader on your Android device. Scroll to the section with online libraries (network libraries) and click on the “Add” button
In addition, more memory and audio output. Optical text recognition programs are very interesting and useful programs. They are used to recognize the script and format it. Often their engines are used to analyze thousands of questionnaires, and their principles are based on pattern recognition.
They certainly have many uses, ranging from replacing human work by checking out large volumes of paper, by transferring library collections into digital content, to helping people with disabilities. Of course, there are other types of apps that you won't hear much about. How do these types of applications perform in the test?
2. In the form that appears, enter “Flibusta” in the “name” and “http://flibusta.net/opds” in the “URL”
2.1 Click on green tick in the upper right corner.
3. In the “network libraries” section, the “Flibusta for website” item (or your option) has appeared, click on it.
Price: free author: Maxim Saplin latest version: 0 Interface language: English Supported languages: English, Bulgarian, Croatian, Czech, Estonian, Danish, French, Spanish, Latin, German, Polish, Portuguese, Romanian, Russian, Serbian, Swedish, Ukrainian, Hungarian, Italian. It's very interesting to be able to recover a document while the program has been paused for a while. The analysis file can only be provided as a graphic file.
And now point by point and with pictures
The configuration options available are limited, but what's interesting is that we can enable things like image or table recognition. The program interface is very poor. When recognizing handwritten text, we must teach our script program ourselves. The analytical document can be provided by scanning or pointing to graphic file. It is worth noting that the material is immediately converted to monochrome, which often resembles a low-quality photocopy. This operation certainly reflects the quality of text recognition.
4. That's all, now you can safely search and download books in Russian directly from CoolReader.
When an error occurs
"error occurred while reading OPDS catalog"
1. Checking the presence of the folder " Books" in the root of the SD card.
(that's right with a capital letter). If the folder does not exist, create it.Recognized material is displayed in text form, where undefined words are listed, which can then be corrected. All that remains is the language that the program must recognize. I was also interested in the recognition times of individual samples by applications, so that was also important. Obviously, this was incomparably small compared to the correctness of the recognized text.
The first test consisted of recognizing a printed document in Polish. The text is recognized flawlessly. Of course, the program had some doubts, but they turned out to be unfounded. Saving text saved. Both programs recognized the text as almost identical.
2. We check that you have entered the URL correctly (OPDS):
http://flibusta.net/opds
You can try the backup directory address:
http://proxy.flibusta.net/opds
Update from 12/01/2015
Using a backup address to access the library
http://flibusta.me/opds
and one more
http://flibusta.is/opds
Tested, works via Beeline Internet.
This is most likely due to the same OCR engine or method. There are many errors in the text in the form of force "pushed" English words, because the program does not have the ability to recognize the Polish language. Also, the text layout is not appropriate.
It is worth adding that the text was, of course, Polish, but in its older form - some of the words were no longer used. Additional obstructions were visible on the next page and the text was slightly slanted. The text is very similar to the original. Only a few words were incorrectly recognized.
The text was also recognized by both programs. The problem that programs have is recognizing the left side of the text due to its slightly higher exposure and its slight slant. The rest of the text was recognized almost flawlessly. Text design It's perfect here. Despite the lack of recognition of the Polish language, the program recognized some text, but, unfortunately, the layout of the text is inappropriate. Some of the words have been recognized as having power on English language. It's a big surprise to recognize the words to the left of the text.
Last time we looked at the issue of reading e-books and got acquainted with the characteristics of one of the best readers(in 6" format).
The question arises: a where can I get these e-books?? Where and how to download them, which libraries are the best, is it possible to download books for free?
The text layout is not suitable. The text was well known in each case. The program also tried to add photos, which it didn't quite succeed. As difficult as it is to guess and recognize this time, it took a little longer. Both programs require quite a lot of time. The program looked for signs in letters and photographs, which didn't work for him. Only one word is recognized well - the rest is a string that has nothing in common.
Not only Polish and English people live here. It was enriched with several photographs. Additionally, the right side was open and the text underneath was visible. There are a few bugs, but they're pretty cosmetic. The text design matches the original. The images are attached correctly, but only what is missing in the line. Interestingly, the program dealt with the more open right side.
In fact, there are a huge number of network libraries, even your eyes run wild when you type the words “ network library" or " e-book library free download».
However, there are the most famous and visited resources where you can conveniently and quickly find and download a book that interests you or a selection of books by one author that interests you. You also need to consider the format in which books are stored there and can be downloaded.
The pictures are also well placed, but there are also lines that are missing. The open side was also recognized. The result of his work again resembles a sequence of random symbols. The first of the extreme tests. The learning test was as follows: someone, trying to recognize the wrong, and gave the English text to recognize a 90° rotation.
The program is "smart enough" that he reversed the text and recognized it flawlessly. For an unacknowledged image, only objections can be made. Here's a round of applause for the developers! As it happens in life - there is no rose without thorns, and the splashes in this test turned out to be at that time. This is incomparably more than other programs, but we must overcome this.
After surfing the net for quite some time and studying the content network libraries, I chose several (not too many) resources for myself, which I began to use successfully. Now I will introduce you to them.
1. I would put the resource first. This huge and constantly growing collection is mainly fiction all genres. This open library, which is replenished by the visitors themselves. All books are available for direct reading online and can be downloaded for free without any restrictions in fb2, epub, mobi and other formats popular for e-books. And even without registration.
The results of both programs are almost identical. Nothing felt right - not even the painting. What the program recognized was an almost random sequence of characters. An attempt was made to recognize fragments of machine text, in Polish, turned towards each other. In addition, the text was saved in different fonts of different sizes.
The test was too demanding for all programs. He rotated the sample so it recognized anything - and recognized only one line of text. It's difficult to comment on OCR times due to poor performance. The third of the more difficult tests. Initially, there was a handwriting recognition test from my laptop, but no program could handle it.
2. Librusec(http://lib.rus.ec) is an old Russian-language library located in the domain of Ecuador. But this library, as far as I remember, was free, but now, unfortunately, it has become commercial and in order to download it requires not only registration, but also purchasing a paid subscription. The cost depends on the validity period of the subscription. For a week – 149 rubles, for a month – 299 rubles, for a year – 2499 rubles. Which, in principle, is a pittance compared to the cost of paper books, considering that, for example, you can download a huge number of books in a week (the number of downloads is not limited).
Once again, the test turned out to be too demanding on the programs. All programs or text fragments are not recognized correctly. Acknowledging the time in the comments is unnecessary because all the programs failed. The last, most demanding test is if no program completes it.
As we can see, the programs could not cope with the sample. Regarding the timing, it is difficult to comment because all the programs are not working correctly. As you can see, computer OCR is as low as it may seem. When there are no puzzles in the text, the probability of error-free recognition is high. However, when the text to be recognized has some difficulties, such as poor text alignment or distortion, the probability of correct recognition decreases.
As bonus I want to tell you about one program for managing your electronic library.
Caliber- a free cross-platform application with open source code for library management and universal e-book converter with big amount opportunities. The program will help you organize your library of books without much difficulty. It also includes: a utility for converting formats, a news aggregator, a program for synchronizing data with electronic device for reading and a good viewer. It is a must-have for fans of reading books on portable mobile devices.
This is a benefit, of course, of the continuous improvement of the product and the fact that it pays. Other programs fall behind in approaching the problem as curiosity rather than as programs vying for position in the commercial market. As for the operating time, it is incomparably shorter than that of its competitor. The program is also poorly specified.
Online e-book resources
Due to frequent changes in access to digital text content of individual organizations at the same home address, please go to home page and start searching for e-books. The Federation of Digital Libraries provides separate data for the year on the dynamics of the project. Currently, there are 113 institutions in the federation, including 105 currently and 6 in preparation. The specified data comes from the official website of the Federation, which also includes detailed statistical materials, tables, graphs, etc.
Library management
Caliber provides the ability to manage your book collection. The concept of the program is a lot of identical books in different formats presented as one book in the library.
You can sort and search books by metadata such as: title, author, publication date, date added, size, book series, rating, etc. There is also support for tags, comments and reviews.
You can easily find the book you need using a simple search or use the advanced search.
Plus Caliber can search online book metadata and covers by title, author or ISBN. The metadata search system is defined by plugins and can be easily extended independently.
Format conversion
Caliber Supports conversion between many e-book formats.
Input formats for conversion supported by the program: CBZ, CBR, CBC, CHM, EPUB, FB2, HTML, LIT, LRF, MOBI, ODT, PDB, PDF, PML, PRC, RB, RTF, TCR, TXT.
Output formats: EPUB, FB2, LIT, LRF, MOBI, OEB, PDB, PDF, PML, RB, RTF, TCR, TXT.
Synchronization with readers
Caliber uses a modular reader driver system that allows you to add new devices. The program supports many devices. When synchronizing Caliber chooses the most suitable format. It also adds all the necessary metadata and information about book series to the reader.
You can download it for free on the developer’s website http://calibre-ebook.com/.
That's all for today.
In order not to miss anything published, you can subscribe to one ofRSS feeds: All site news, News of the “Tales of the Old User” section or News of the “Android World” section.
This article can be discussed at