tesseract-ocr | package contains the Tesseract Open Source OCR Engine | Computer Vision library

The cv2 pre-processing is unecessary here, tesseract is able to find the text on its own. See the example below, commented inline:

There are several tutorial how to build tesseract on windows with cmake and VS e.g. https://bucket401.blogspot.com/2021/03/building-tesserocr-on-ms-windows-64bit.html (you can ignore end of tutorial - python module), minimalist tesseract or with clang

I solved the error by giving the full path of the image in the python script to imread()

[EDIT]

For anyone wondering, the image in the question was updated after posting my answer. That was the original image:

Thus, the below output in my original answer.

That's the newly posted image:

The specific Turkish characters, especially in the last word, are still not properly detected (since I still can't use lang='tur' right now), but at least the Ö and Ü can be detected using lang='deu', which I have installed:

Here's one possible solution. This is in Python, but it should be clear enough for a Java port. We will apply a method called gained division. The idea is that you try to build a model of the background and then weight each input pixel by that model. The output gain should be relatively constant during most of the image. This will get rid of most of the background color variation. We can use a morphological chain to clean the result a little bit, let's see the code:

Have a look at the page segmentation modes of pytesseract, cf. this Q&A. For example, using config='-psm 12' will already give you all desired texts. Nevertheless, those graphs are also somehow interpreted as texts.

That's why I would preprocess the image to get single boxes (actual texts, the graphs, those information from the top, etc.), and filter to only store those boxes with the content of interest. That could be done by using

• the y coordinate of the bounding rectangle (not in the upper 5 % of the image, that's the mobile phone status bar),
• the width w of the bounding rectangle (not wider than 50 % of the image' width, these are the horizontal lines),
• the x coordinate of the bounding rectangle (not in middle third of the image, these are the graphs).

What's left is to run pytesseract on each cropped image with config='-psm 6' for example (assume a single uniform block of text), and clean the texts from any line breaks.

That'd be my code:

Notice how your power meters either use blue or green LEDs to light up the display; I suggest you use this color display to your advantage. What I'd do is select only one RGB channel based on the LED color. Then I can threshold it based on some algorithm or assumption. After that, you can do the downstream steps of cropping / resizing / transformation / OCR etc.

For example, using your example image 1, look at its histogram here. Notice how there is a small peak of green to the right of the 150 mark.

I take advantage of this, and set anything below 150 to zero. My assumption being that the green peak is the bright green LED in the image.

This link provides me the answer. Its removing the noise in the background image.

The underlying API that watchdog uses to monitor linux filesystem events is called inotify. The Docker for Windows WSL 2 backend documentation notes:

Linux containers only receive file change events (“inotify events”) if the original files are stored in the Linux filesystem.

The directory you're mounting, c:\My_MR, resides on the Windows file system and thus inotify inside the watcher container doesn't work.

Instead, you can run docker from inside your WSL 2 default distribution with a linux filesystem path, e.g., ~/my_mr: