tessdata icon indicating copy to clipboard operation
tessdata copied to clipboard

When using jpn.traineddata, numbers are enclosed by cricle

Open atuyosi opened this issue 6 years ago • 1 comments

I got strange output with tessdata/jpn.traineddata.

sample image is below.

digit-sample

$ export TESSDATA_PREFIX=/path/to/tessdata
$ tesseract -l eng digit-sample.png  stdout
Warning. Invalid resolution 0 dpi. Using 70 instead.
Estimating resolution as 172
0123456789

$ tesseract -l jpn digit-sample.png  stdout
Warning. Invalid resolution 0 dpi. Using 70 instead.
Estimating resolution as 172
⑥①②③④⑤⑥⑦⑧⑨

$ tesseract -l jpn  digit-sample.png stdout --oem 0
Warning. Invalid resolution 0 dpi. Using 70 instead.
Estimating resolution as 172
0ー23456789

$ tesseract -l jpn  digit-sample.png stdout --oem 1
Warning. Invalid resolution 0 dpi. Using 70 instead.
Estimating resolution as 172
⑥①②③④⑤⑥⑦⑧⑨

But tessdata/Japanese.traineddata', testdata_fast/jpn.traineddataandtessdata_best/jpn.traineddata` are no problem.

I think that tessdata/jpn.traineddata has some wrong data.

Environment information:

tesseract 4.0.0-beta.4-20-ge9b4e
 leptonica-1.76.0
  libjpeg 9c : libpng 1.6.35 : libtiff 4.0.9 : zlib 1.2.11
 Found AVX
 Found SSE

atuyosi avatar Aug 05 '18 13:08 atuyosi

ParamsModel::Incomplete line � ParamsModel::Incomplete line � ParamsModel::Incomplete line ParamsModel::Incomplete line �� ParamsModel::Incomplete line ParamsModel::Incomplete line ParamsModel::Incomplete line =� ParamsModel::Incomplete line ParamsModel::Incomplete line SJ� ParamsModel::Incomplete line � ParamsModel::Incomplete line ParamsModel::Incomplete line �\� ParamsModel::Incomplete line ParamsModel::Incomplete line ParamsModel::Incomplete line P� ParamsModel::Incomplete line ParamsModel::Incomplete line `<E ParamsModel::Incomplete line � ParamsModel::Incomplete line ParamsModel::Incomplete line �,� ParamsModel::Incomplete line ParamsModel::Incomplete line jpn:synth20170629:[1,48,0,1Ct3,3,16Mp3,3Lfys64Lfx96Lrx96Lfx512O1c1] 2019-08-13 02:53:18.783 INFO 20928 --- [lication Thread] com.xhxj.ocr.controller.MainController : ”ユウト : ま、 待ってよ。 写真く らいいく らでも撮るから引っ張らないで。

I have errors every time I run

XHXJ avatar Aug 12 '19 18:08 XHXJ