Problem solved by additional preprocessing - blending with white background. This form of SetImage(): SetImage(avframe->data[0], avframe->width, avframe->height, 4, avframe->linesize[0]); do not removing alpha, so I need to do it myself. On Thursday, December 23, 2021 at 11:37:27 PM UTC+3 Yesbird wrote:
> And sorry, language initialization from API is the same as from comman > line: > std::string lang = "ron"; > > On Thursday, December 23, 2021 at 11:28:18 PM UTC+3 Yesbird wrote: > >> Hi, guys ! >> >> I am doing video subtitles recognition for one of my C++ projects and can >> not figure out why for the same image tesseract gives good results when I >> run it from command line but fails from API. I see a couple of different >> parameters when running >> >> tesseract --print-parameters >> >> and don't know how to find which of them affect results. >> >> Could anyone help me, please ? >> >> -- From command line ---------- >> tesseract ./subtitles/sub_ron_1.png stdout -l ron --dpi 600 >> ----------------------------------------- >> Turul virtual făcut de Kira şi Matt >> a fost foarte amuzant. >> ----------------------------------------- >> >> -- From API ------------------------- >> char *text; >> std::string lang = "rum"; >> ocr->Init(NULL, lang.c_str()); >> ocr->SetImage(avframe->data[0], avframe->width, avframe->height, 4, >> avframe->linesize[0]); >> text = ocr->GetUTF8Text(); >> ----------------------------------------- >> € II a] e E ăn si 2 W a p:] VA >> Turul'virtual făcut de Kira şi Matt >> nat fn arte - SE, >> a fost foarte amuzant. >> ----------------------------------------- >> >> -- Version info --------------------------- >> tesseract 4.1.1 >> leptonica-1.79.0 >> libgif 5.1.4 : libjpeg 8d (libjpeg-turbo 2.0.3) : libpng 1.6.37 : libtiff >> 4.1.0 : zlib 1.2.11 : libwebp 0.6.1 : libopenjp2 2.3.1 >> Found SSE >> Found libarchive 3.4.0 zlib/1.2.11 liblzma/5.2.4 bz2lib/1.0.8 >> liblz4/1.9.2 libzstd/1.4.4 >> --------------------------------------------- >> >> Image: >> [image: sub_ron_1.png] >> > -- You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/6ab63613-421c-42c8-8efb-e2c4390e5228n%40googlegroups.com.
