summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorNick White <git@njw.name>2021-02-09 17:55:18 +0000
committerNick White <git@njw.name>2021-02-09 17:55:18 +0000
commitd134ae8a3be1ddc79122937609fa441a5076fd03 (patch)
treee0986995d5f6bbcc4a4e13d618dfddff2815fe72
parent45943f847b3db8db5142c79a806f251659264ca0 (diff)
Add extracthocrlines tool
-rw-r--r--cmd/extracthocrlines/main.go89
1 files changed, 89 insertions, 0 deletions
diff --git a/cmd/extracthocrlines/main.go b/cmd/extracthocrlines/main.go
new file mode 100644
index 0000000..d765875
--- /dev/null
+++ b/cmd/extracthocrlines/main.go
@@ -0,0 +1,89 @@
+// Copyright 2021 Nick White.
+// Use of this source code is governed by the GPLv3
+// license that can be found in the LICENSE file.
+
+// extracthocrlines copies the text and corresponding image section
+// for each line of a HOCR file into separate files, which is
+// useful for OCR training
+package main
+
+import (
+ "flag"
+ "fmt"
+ "io"
+ "log"
+ "os"
+ "path/filepath"
+
+ "rescribe.xyz/utils/pkg/hocr"
+ "rescribe.xyz/utils/pkg/line"
+)
+
+const usage = `Usage: extracthocrlines file.hocr [file.hocr]
+
+Copies the text and corresponding image section for each line
+of a HOCR file into separate files, which is useful for OCR
+training.
+`
+
+// saveline saves the text and image for a line in a directory
+func saveline(l line.Detail, dir string) error {
+ err := os.MkdirAll(dir, 0700)
+ if err != nil {
+ return err
+ }
+
+ base := filepath.Join(dir, l.OcrName+"_"+l.Name)
+
+ f, err := os.Create(base + ".png")
+ if err != nil {
+ return fmt.Errorf("Error creating file %s: %v", base+".png", err)
+ }
+
+ err = l.Img.CopyLineTo(f)
+ if err != nil {
+ return fmt.Errorf("Error writing line image for %s: %v", base+".png", err)
+ }
+
+ f, err = os.Create(base + ".txt")
+ if err != nil {
+ return fmt.Errorf("Error creating file %s: %v", base+".txt", err)
+ }
+
+ _, err = io.WriteString(f, l.Text)
+ if err != nil {
+ return fmt.Errorf("Error writing line text for %s: %v", base+".txt", err)
+ }
+
+ return nil
+}
+
+func main() {
+ flag.Usage = func() {
+ fmt.Fprintf(flag.CommandLine.Output(), usage)
+ flag.PrintDefaults()
+ }
+ dir := flag.String("d", ".", "Directory to save lines in")
+ flag.Parse()
+ if flag.NArg() < 1 {
+ flag.Usage()
+ os.Exit(1)
+ }
+
+ for _, f := range flag.Args() {
+ newlines, err := hocr.GetLineDetails(f)
+ if err != nil {
+ log.Fatal(err)
+ }
+
+ for _, l := range newlines {
+ if l.Img == nil {
+ continue
+ }
+ err = saveline(l, *dir)
+ if err != nil {
+ log.Fatal(err)
+ }
+ }
+ }
+}