improve ocropy processor (#8)

* fix requirements

* improve ocrd-cis-ocropy-recognize:

- add proper error handling
- use proper temporary files
- re-introduce binarization (was commented)
- add sanity checks (from ocropy CLI)
- make de-warping optional

* improve ocrd-cis-ocropy-recognize again:

- abolish temporary files altogether:
  keep converting between pillow and array
  formats in memory
- make logger available to all functions
- make binarization method and dewarping
  selectable via ocrd-tool parameters
- add binarization method from original
  ocropus-nlbin (including local whitelevel
  estimation and de-skewing)
- calculate OCR-GT distances while processing
  and show CER per input file

ocrd_cis/ocrd-tool.json

@@ -56,6 +56,16 @@
 			],
 			"description": "Recognize text in lines with ocropy",
 			"parameters": {
+                                "dewarping": {
+                                    "type": "boolean",
+                                    "description": "enable line normalization",
+                                    "default": true
+                                },
+                                "binarization": {
+                                    "type": "string",
+                                    "enum": ["none", "global", "otsu", "gauss-otsu", "ocropy"],
+                                    "default": "none"
+                                },
 				"textequiv_level": {
 					"type": "string",
 					"enum": ["line", "word", "glyph"],

ocrd_cis/ocropy/ocrolib/common.py

@@ -139,7 +139,7 @@ def array2pil(a):
         else:
             raise OcropusException("bad image rank")
     elif a.dtype==dtype('float32'):
-        return PIL.Image.fromstring("F",(a.shape[1],a.shape[0]),a.tostring())
+        return PIL.Image.frombytes("F",(a.shape[1],a.shape[0]),a.tostring())
     else:
         raise OcropusException("unknown image type")