.NET中两种OCR方式对比

PRivate static void TesseractSample()        {            var testImagePath = "phototest.tif";            try            {                using (var engine = new TesseractEngine(@"../../tessdata", "eng", EngineMode.Default))                {                    using (var img = Pix.LoadFromFile(testImagePath))                    {                        using (var page = engine.Process(img))                        {                            var text = page.GetText();                            Console.WriteLine("Mean confidence: {0}", page.GetMeanConfidence());                            Console.WriteLine("Text (GetText): /r/n{0}", text);                            Console.WriteLine("Text (iterator):");                            using (var iter = page.GetIterator())                            {                                iter.Begin();                                do                                {                                    do                                    {                                        do                                        {                                            do                                            {                                                if (iter.IsAtBeginningOf(PageIteratorLevel.Block))                                                {                                                    Console.WriteLine("<BLOCK>");                                                }                                                Console.Write(iter.GetText(PageIteratorLevel.Word));                                                Console.Write(" ");                                                if (iter.IsAtFinalOf(PageIteratorLevel.TextLine, PageIteratorLevel.Word))                                                {                                                    Console.WriteLine();                                                }                                            } while (iter.Next(PageIteratorLevel.TextLine, PageIteratorLevel.Word));                                            if (iter.IsAtFinalOf(PageIteratorLevel.Para, PageIteratorLevel.TextLine))                                            {                                                Console.WriteLine();                                            }                                        } while (iter.Next(PageIteratorLevel.Para, PageIteratorLevel.TextLine));                                    } while (iter.Next(PageIteratorLevel.Block, PageIteratorLevel.Para));                                } while (iter.Next(PageIteratorLevel.Block));                            }                        }                    }                }            }            catch (Exception e)            {                Console.WriteLine("Unexpected Error: " + e.Message);                Console.WriteLine("Details: ");                Console.WriteLine(e.ToString());            }            Console.WriteLine("Press anykey to end");            Console.ReadKey();        }使用MS cognitive做OCR1. 有一个Cognitive 的账号。2. 准备一个测试图片
3. 测试代码：
... MS_OCRApi("phototest.tif");...    static async void MS_OCRApi(string pathOfImage)        {                        var client = new HttpClient();            var queryString = HttpUtility.ParseQueryString(string.Empty);            const string API_KEY = "{your key}";            // Request headers            client.DefaultRequestHeaders.Add("Ocp-Apim-Subscription-Key", API_KEY);            // Request parameters            queryString["language"] = "unk";            queryString["detectOrientation "] = "true";            var uri = "https://westus.api.cognitive.microsoft.com/vision/v1.0/ocr?" + queryString;            HttpResponseMessage response;            // Request body            byte[] byteData = File.ReadAllBytes(pathOfImage);            using (var content = new ByteArrayContent(byteData))            {                content.Headers.ContentType =                    new MediaTypeHeaderValue("application/octet-stream");                response = await client.PostAsync(uri, content);                var s = new MemoryStream();                await response.Content.CopyToAsync(s);                var str = Encoding.UTF8.GetString(s.ToArray());                var results = JsonConvert.DeserializeObject<MsOCRResult>(str);                Console.WriteLine("Results :");                var lines = results.regions.SelectMany(x => x.lines);                foreach (var ocrLine in lines)                {                    Console.WriteLine(ocrLine);                }                Console.ReadKey();            }        }
对比两种OCR：1. MS Cognitive Service是RESTful api based，支持多种客户端而且不用关心机器学习过程，但是收费(每月5000个call，1分钟20个call免费)。不过可考虑使用缓存来降低成本。2. Tesseract是dll based。因此需要不断更新训练数据，如果需要支持多客户端，需要自己wrap一个RESTful。好处就是免费。