Changes as proposed in #14 #15
Conversation
Makefile
Outdated
| @@ -3,15 +3,22 @@ export | |||
| SHELL := /bin/bash | |||
| LOCAL := $(PWD)/usr | |||
| PATH := $(LOCAL)/bin:$(PATH) | |||
| TESSDATA = $(LOCAL)/share/tessdata | |||
| LANGDATA = $(PWD)/langdata-$(LANGDATA_VERSION) | |||
| HOME := /home/ubuntu | |||
There was a problem hiding this comment.
$HOME is set by default (i.e. HOME := $(HOME)), if you want to override it you can, but I would not set a different default. As you use it, it seems more like you meant PWD or REPO_ROOT?
Makefile
Outdated
| LANGDATA = $(PWD)/langdata-$(LANGDATA_VERSION) | ||
| HOME := /home/ubuntu | ||
| TESSDATA = $(HOME)/tessdata_best | ||
| LANGDATA = $(HOME)/langdata |
There was a problem hiding this comment.
I find it preferable to have the explicit version of langdata for reproducible results.
Makefile
Outdated
|
|
||
| # Name of the model to be built | ||
| MODEL_NAME = foo | ||
|
|
||
| # Name of the model to continue from | ||
| CONTINUE_FROM = frk |
There was a problem hiding this comment.
Default to $(MODEL_NAME)?
Makefile
Outdated
|
|
||
| $(ALL_LSTMF): $(sort $(patsubst %.tif,%.lstmf,$(wildcard $(TRAIN)/*.tif))) | ||
| find $(TRAIN) -name '*.lstmf' -exec echo {} \; | sort -R -o "$@" | ||
|
|
||
| $(TRAIN)/%.lstmf: $(TRAIN)/%.box | ||
| tesseract $(TRAIN)/$*.tif $(TRAIN)/$* lstm.train | ||
| tesseract $(TRAIN)/$*.tif $(TRAIN)/$* --psm 6 lstm.train |
There was a problem hiding this comment.
Make page segmentation mode configurable?
|
I had changed it for my own testing. Please make the required changes for the generic script. Some related discussion is at https://groups.google.com/forum/?utm_medium=email&utm_source=footer#!msg/tesseract-ocr/be4-rjvY2tQ/vIF2a0XbCgAJ |
There was a problem hiding this comment.
Nice. I am excited to test CONTINUE_FROM. Many thanks to @kba and @Shreeshrii
Thanks for spotting, that was a mistake, fixed in master. |
@Shreeshrii @wrznr