Merge pull request #1 from arthurhenrique/test-perf

chore(*): performance test

Dockerfile

@@ -1,4 +1,4 @@
-FROM golang:1.15.3-alpine3.12
+FROM golang:1.15.5-alpine3.12
 RUN apk add --update --no-cache --virtual wget-dependencies \
     ca-certificates \
     openssl \

README.md

@@ -2,6 +2,9 @@
 
 POC to test lib tesseract with portuguese and english language.
 
+edit: now just portuguese for performance:
+[test perf](https://github.com/arthurhenrique/go-ocr/pull/1)
+
 ## Notes
 
 This works properly using images with 300 dpi <= x <= 600 dpi.

convert.go

@@ -2,14 +2,11 @@ package main
 
 import (
 	"bufio"
-	"bytes"
 	"fmt"
-	"image"
-	"image/color"
-	"image/jpeg"
 	"os"
 )
 
+// convertToBytes return bytes
 func convertToBytes(fileName string) []byte {
 	file, err := os.Open(fileName)
 	if err != nil {

detection.go

@@ -2,13 +2,11 @@ package main
 
 import (
 	"strings"
-	"sync"
 	"unicode"
 
 	"github.com/otiai10/gosseract/v2"
 )
 
-var once sync.Once
 var instance *gosseract.Client
 
 // TextMethod contains configuration of tesseract client

main.go

@@ -2,38 +2,54 @@ package main
 
 import (
 	"fmt"
+	"log"
+	"time"
 
 	"github.com/otiai10/gosseract/v2"
 )
 
 func main() {
-
-	fileName := "files/2.jpg"
-
-	instance = gosseract.NewClient()
-
-	tm := TextMethod{
-		Name:     "tesseract",
-		Language: "por+eng",
-		Variables: map[string]string{
-			"tessedit_pageseg_mode": "3", // auto page segmentation mode
-			"load_system_dawg":      "0", // removing dict to increase recognition
-			"load_freq_dawg":        "0",
-		},
-		Client: instance,
-	}
-	defer instance.Close()
-
-	// Handler tesseract's settings
-	tm.tesseractSettings()
-
-	bytesImage := convertToBytes(fileName)
-	object, err := tm.extract(bytesImage)
-	if err != nil {
-		return
-	}
-
-	if object != nil {
-		fmt.Print(*object)
+	iterations := 50
+	total := time.Duration(0)
+
+	for i := 0; i < iterations; i++ {
+		start := time.Now()
+
+		fileName := "files/bode.jpg"
+
+		// fast model
+		// https://github.com/tesseract-ocr/tessdata_fast/raw/master/por.tra
+		instance = gosseract.NewClient()
+
+		tm := TextMethod{
+			Name:     "tesseract",
+			Language: "por",
+			Variables: map[string]string{
+				"tessedit_pageseg_mode": "3", // auto page segmentation mode
+				"load_system_dawg":      "0", // removing dict to increase recognition
+				"load_freq_dawg":        "0",
+			},
+			Client: instance,
+		}
+		defer instance.Close()
+
+		// Handler tesseract's settings
+		tm.tesseractSettings()
+
+		bytesImage := convertToBytes(fileName)
+		object, err := tm.extract(bytesImage)
+		if err != nil {
+			return
+		}
+
+		if object != nil {
+			fmt.Println(*object)
+		}
+		elapsed := time.Since(start)
+		log.Printf("took %s", elapsed)
+		total += elapsed
 	}
+	log.Printf("All Duration: %s", total)
+	log.Printf("loops: %d", iterations)
+	log.Printf("Average: %s", total/time.Duration(iterations))
 }