public async Task ScanPdf(string pdfFile, string range, string scanFile)
{
OcrRequest request = CreateOcrRequest();
request.DocumentFile = pdfFile;
request.PageRange = range;
OcrResult <OcrProcessDocumentResponse> response = await _ocrWebService.ProcessDocument(request);
if (response.Success)
{
//string directory = GetScanDirectory();
//zdir.CreateDirectory(scanDirectory);
zfile.CreateFileDirectory(scanFile);
//string scanFile = zPath.Combine(scanDirectory, "scan");
Trace.WriteLine($"save scan to \"{scanFile}\"");
await _ocrWebService.DownloadResultDocuments(response.Data, scanFile);
Trace.WriteLine($"scan ok {response.Data.ProcessedPages} pages - remainder {response.Data.AvailablePages} pages");
}
else
{
Trace.WriteLine($"scan error code {response.StatusCode}");
}
}
/// <summary>
/// Runs optical character recognition on the given image.
/// </summary>
/// <param name="imageUrl">URL to the image</param>
/// <returns>The OCR result</returns>
public async Task <OcrResult> RunOcrAsync(string imageUrl)
{
var url = $"https://{endpoint}/{OcrPath}?detectOrientation=true";
var request = new HttpRequestMessage(HttpMethod.Post, url);
request.Headers.Add(SubscriptionKeyHeaderName, this.accessKey);
var body = new OcrRequest {
Url = imageUrl
};
request.Content = new ObjectContent <OcrRequest>(body, JsonFormatter);
var response = await this.httpClient.SendAsync(request);
var result = await this.ProcessApiResponse <OcrResult>(response);
return(result);
}
public async Task ScanImage(string imageFile, string scanFile)
{
OcrRequest request = CreateOcrRequest();
request.DocumentFile = imageFile;
OcrResult <OcrProcessDocumentResponse> response = await _ocrWebService.ProcessDocument(request);
if (response.Success)
{
//string scanDirectory = GetScanDirectory();
//zdir.CreateDirectory(scanDirectory);
zfile.CreateFileDirectory(scanFile);
//string scanFile = zPath.Combine(scanDirectory, $"scan-page-{page:000}");
Trace.WriteLine($"save scan to \"{scanFile}\"");
await _ocrWebService.DownloadResultDocuments(response.Data, scanFile);
Trace.WriteLine($"scan ok {response.Data.ProcessedPages} pages - remainder {response.Data.AvailablePages} pages");
}
}
public static Result <string> ExtractWordFromImage(OcrRequest request)
{
Image myImage = WindowsHelper.CaptureWindowByName(request.ProcessName, new Rectangle(request.ImageXCoordinate, request.ImageYCoordinate, request.ImageWidth, request.ImageHeight));
var imageAttr = new ImageAttributes();
imageAttr.SetThreshold(0.75f);
var bm = new Bitmap(myImage.Width, myImage.Height);
using (var gr = Graphics.FromImage(bm))
{
gr.DrawImage(myImage, new[] {
new Point(0, 0),
new Point(myImage.Width, 0),
new Point(0, myImage.Height),
}, new Rectangle(0, 0, myImage.Width, myImage.Height),
GraphicsUnit.Pixel, imageAttr);
}
if (request.InvertColour)
{
for (var h = 0; h < bm.Height; h++)
{
for (var w = 0; w < bm.Width; w++)
{
var color = bm.GetPixel(w, h);
bm.SetPixel(w, h, Color.FromArgb(255, (255 - color.R), (255 - color.G), (255 - color.B)));
}
}
}
var fp = $"{Path.GetTempPath()}{Guid.NewGuid()}.jpg";
bm.Save(fp);
var word = new TesseractEngine($"{new System.IO.DirectoryInfo(Environment.CurrentDirectory).FullName}\\tessdata\\",
"eng",
EngineMode.Default).Process(Pix.LoadFromFile(@"D:\\tmp.jpg"))
.GetText()
.Replace("\n", " ");
return(Result.SuccessIf <string>(!string.IsNullOrWhiteSpace(word), word, $"No words readable, image saved in {fp}"));
}
