rmast

This might be the fix for the upper disappearing word of issue 3871

3

https://github.com/tesseract-ocr/tesseract/issues/3871

Ghost block?

9

Despite all effort in the inverted text [my good quality proof-picture](https://user-images.githubusercontent.com/3341558/179373903-ef6cc246-f4e5-4633-a762-ded4dd22708f.jpg) is still not correctly segmented and read. The clear to be read for a human sentence near the bottom,...

I want specific non-fatal messages to precede others and interpret long sentences.

1

The newest version I found still can't cope with lines that have been split up by different rules of the max. linesize, like The mail system <xxxxxxxxxxxxxxxxxxxxxx>: maildir delivery failed:...

python mnist.py --device ocl:0 runs on UHD Graphics 630, remains heavy load without progress after first epoch

5

The test with the following statement shows one epoch on the screen, and then seems to keep the GPU and CPU running whithout any further progress on the screen for...

pixAutoPhotoinvert - the sequel

2

 results with defaults 0,NULL,NULL in  The KVK on top left should be left as it was, or at least the bottom half of the V should be...

### Bug description When I make the detector detect text in the following image  the preferred dots in the boxes are from the i's in the...

Why should I be forced to have a CUDA or ROCm machine when wanting to run OpenVino on Intel?

2

This link tells me ort-inference supports OpenVino: https://github.com/pytorch/ort#-inference "ONNX Runtime for PyTorch supports PyTorch model inference using ONNX Runtime and Intel® OpenVINO™. It is available via the torch-ort-infer python package....

pdfcomp: problems with inverted text that is often better in hocr.

10

This form [https://www.kvk.nl/download/Formulier-14-wijziging-ondernemings-en-vestigingsgegevens_tcm109-365607.pdf](url) First page saved to jpeg via this site: https://smallpdf.com  Result of the left column is quite readable at the right screen-resolution. ``` ocrmypdf --pdfa-image-compression lossless -O0...

correct ratio determination for noise estimation

5

I solved issue https://github.com/internetarchive/archive-pdf-tools/issues/52 myself.

Bug in foreground/background separator choosing massive block instead of character outline.

14

Partly anonymized replay of my previous finding on compressing the bankstatement with downsampling the foreground, revealing a bug in the foreground-binarizer/separator.  Add fg_downsample=12 in compress-pdf-images: mrc_gen = create_mrc_hocr_components(pil_image, hocr_word_data,...