Skip to content

Latest commit

 

History

History
124 lines (108 loc) · 4.68 KB

README.md

File metadata and controls

124 lines (108 loc) · 4.68 KB

Image-Processing

  • Image Processing (Windows)
  • C/C++ development tool for image processing, computer vision and machine learning

Introduction

  • Open program sources (without OpenCV)
  • Visual Studio 2019, 2017, 2015, 2013 and lower
  • GUI: MFC
  • Win32 and x64

Source codes

  • Bitmap load and save (cpp), jpeg load and save (lib)
  • Format conversion (RGB, YUV420, YUV422, YUV444)
  • Low level image processing (spatial filtering (including fast versions), frequency transform (DFT, FFT, DCT), ...)
  • Geometric transformation (affine, bilinear, perspective, ...)
  • Simple operations (Hough, boundary following, labeling, thinning, morphological operations, ...)
  • Drawing functions for text (Windows), line, circle and ellipse
  • ROI setting (AABB and polygons)
  • Web camera processing
  • Image sequence saving
  • Easy debugging (image, text, listbox)
  • Sequential processing
    • Image sequences (bmp, jpeg, pcx, ppm, png, raw)
    • Avi sequences
    • YUV sequences

  • Multi-layer perceptron (MLP)
  • Avi recording

Example code for sequential processing

  1. Add a new class
  2. Add a new processing method (1D array image - BGR, 24bit, 4byte aligned)
  3. Add an instance member to CMainProcessDlg class (MainProcessDlg.h)
  4. Call the processing method as follows:
void CMainProcessDlg::OnTimer(UINT_PTR nIDEvent) {
    ...
			else if(nIDEvent == TID_SEQUENCE_RUN) {
				Sleep(m_nSleep);
				...
				else if(FileName.Right(4) == _T(".JPG")) {
					Image1D = ReadJpeg(mbcsFileName, &nW, &nH);
				}

				if(Image1D) {
					// Processing
					DisplayCimage1D(Image1D, nW, nH, 0, 0, false, true);
					// Instance.NewMethodExample(Image1D, nW, nH); // call 
					//
					m_bFirstFrame = false;
					delete [] Image1D;
				}
			
				CString DispFrameNum;
				DispFrameNum.Format(_T("%d(0,%d)"), m_nCurrentProcessingNum, m_nSequenceLength-1);
				SetDlgItemText(IDC_TOTAL_FRAME, DispFrameNum);

				m_nCurrentProcessingNum++;
			}
  ...
}

Example code for connected-component labeling

int **Label = imatrix(nH, nW);
int SmallThre = 20; 	// the minimum area threshold for eliminating small connected components

// Image: 	foregrounds	ImageGray[y][x] > 128, 
// 		backgrounds 	otherwise
int LabelCnt = Labeling(Image, Label, nW, nH, SmallThre); 	// returns the number of labels
								// Labels are started from 0
								// Background regions are labeled as -1

if(LabelCnt > 0)
{
	CIpRect *pBoundBox = new CIpRect[LabelCnt]; 		// Bounding box (AABB: axis aligned bounding box)
	SetLabelBoundBox(Label, nW, nH, pBoundBox, LabelCnt); 	// extracts bounding boxes
	
	// Bounding box processing
	
	delete [] pBoundBox;
}
		
free_imatrix(Label, nH, nW);

Example code for displaying images

unsigned char **ImageGray;	// 2D image array (grayscale)
unsigned char *Image1D; 	// 24bit color, BGR, 4byte alligned
unsigned char **ImageRed, **ImageGreen, **ImageBlue; // 2D image array (red, green and blue components)
int **IIimage2D;		// 2D image array (int)
double **DImage2D;		// 2D image array (double)

int nW, nH;		// Image size
int nPosX, nPosY; 	// Display position
int Scale = 100;	// Display scale

// bool DisplayCimage2D(unsigned char  **ImageGray, int nW, int nH, int nPosX, int nPosY, 
//	bool bErase = true, bool bDelete = false, int Rate = 100);
// if bErase is ture, the images which are not listed are erased
// if bDelete is ture, the previously displayed image having the same size and the same position 
//			with the current image is deleted from the dispaly list
DisplayCimage2D(ImageGray, nW, nH, nPosX, nPosY, false, true, Scale);
DisplayIimage2D(IIimage2D, nW, nH, nPosX, nPosY, false, true);	// IIimage2D data are normalized to [0, 255]
DisplayDimage2D(DImage2D, nW, nH, nPosX, nPosY, false, true);	// DImage2D data are normalized to [0, 255]
DisplayCimage1D(Image1D, nW, nH, nPosX, nPosY, false, true, Scale);
DisplayCimage2DColor(ImageRed, ImageGreen, ImageBlue, nW, nH, nPosX, nPosY, false, true, Scale);

Visual Studio

  • Visual Studio 2019, 2017, 2015, 2013 and lower
    • MFC libraries are required
    • Project setting
  • Visual Studio 2013 and lower
    • Replace JpegLib.lib, JpegLibD.lib, JpegLib_64.lib and JpegLibD_64.lib with those in Libs(2013)
  • If the MFC libraries were not installed
    1. Run 'Visual Studio Installer' by selecting the menu: [Tools]-[Get Tools and Features...]
    2. Check [Desktop development with C++] in Workloads tab
    3. Check [MFC and ATL support(x86 and x64)] in Summary
    4. Click [Modify] button