I am using scrcpy to synchronize the screen of my old Android phone to the computer device. This way I can monitor my phone screen on the computer. Everyone knows that scrcpy can set the screen compression in order to transmit the screen smoothly. Now I need to automatically recognize some "Signature marker" on the screen, and then get the notification...
For example: I have an app on my old mobile phone. When the app gets the alarm message, it can pop up a specific image. I want to compare the image to identify whether the message has been received. I found a lot of color contrast methods in the forum. For example, the replacement color code by dilettante, this can be obtained if the appropriate tolerance is set. Of course, the success rate depends on the difference between the pattern and the background color.
It's like identifying an apple in a picture. The color of this apple may be slightly different every time. This is because the screen of the mobile phone is compressed during the scrcpy transmission process.
Last edited by xxdoc123; Dec 9th, 2021 at 10:01 PM.
VERSION 5.00
Begin VB.Form Form1
Caption = "Form1"
ClientHeight = 6168
ClientLeft = 108
ClientTop = 456
ClientWidth = 10128
BeginProperty Font
Name = "Tahoma"
Size = 7.8
Charset = 0
Weight = 400
Underline = 0 'False
Italic = 0 'False
Strikethrough = 0 'False
EndProperty
LinkTopic = "Form1"
ScaleHeight = 514
ScaleMode = 3 'Pixel
ScaleWidth = 844
ShowInTaskbar = 0 'False
StartUpPosition = 3 '窗口缺省
Begin VB.PictureBox Picture3
AutoSize = -1 'True
Height = 4212
Left = 3960
ScaleHeight = 4164
ScaleWidth = 4848
TabIndex = 3
TabStop = 0 'False
Top = 1440
Width = 4896
End
Begin VB.CommandButton Command1
Caption = "Command1"
Height = 360
Left = 4560
TabIndex = 2
Top = 0
Width = 990
End
Begin VB.PictureBox Picture2
AutoSize = -1 'True
Height = 684
Left = 6000
Picture = "Form1.frx":0000
ScaleHeight = 636
ScaleWidth = 780
TabIndex = 1
TabStop = 0 'False
Top = 720
Width = 828
End
Begin VB.PictureBox Picture1
AutoSize = -1 'True
Height = 2316
Left = 120
Picture = "Form1.frx":28D6
ScaleHeight = 2268
ScaleWidth = 2952
TabIndex = 0
TabStop = 0 'False
Top = 240
Width = 3000
End
End
Attribute VB_Name = "Form1"
Attribute VB_GlobalNameSpace = False
Attribute VB_Creatable = False
Attribute VB_PredeclaredId = True
Attribute VB_Exposed = False
Option Explicit
Private Sub Command1_Click()
Picture1.ScaleMode = 3
Picture1.AutoRedraw = True
Picture1.AutoSize = True
Picture2.ScaleMode = 3
Picture2.AutoRedraw = True
Picture2.AutoSize = True
Picture3.Width = Picture1.Width
Picture3.Height = Picture1.Height
Picture3.AutoRedraw = True
Picture3.AutoSize = True
If Picture2.ScaleWidth > Picture1.ScaleWidth Or Picture2.ScaleHeight > Picture1.ScaleHeight Then
MsgBox "The image size does not meet the requirements"
Exit Sub
End If
Command1.Enabled = False
Call ReadPic(Picture1.Image, iDATA())
Call ReadPic(Picture2.Image, bDATA())
Call TemplateMatchDib(Picture1.ScaleWidth, Picture1.ScaleHeight, Picture2.ScaleHeight, Picture2.ScaleWidth)
Command1.Enabled = True
Call ShowPic(Picture3.Image, iDATA)
End Sub
Code:
Attribute VB_Name = "Module1"
Option Explicit
Public Declare Function StretchBlt _
Lib "gdi32" (ByVal hdc As Long, _
ByVal X As Long, _
ByVal y As Long, _
ByVal nWidth As Long, _
ByVal nHeight As Long, _
ByVal hSrcDC As Long, _
ByVal XSrc As Long, _
ByVal YSrc As Long, _
ByVal nSrcWidth As Long, _
ByVal nSrcHeight As Long, _
ByVal dwRop As Long) As Long
Private Declare Function GetDIBits _
Lib "gdi32" (ByVal aHDC As Long, _
ByVal hBitmap As Long, _
ByVal nStartScan As Long, _
ByVal nNumScans As Long, _
lpBits As Any, _
lpBI As BITMAPINFO, _
ByVal wUsage As Long) As Long
Private Declare Function CreateCompatibleDC Lib "gdi32" (ByVal hdc As Long) As Long
Private Declare Function DeleteDC Lib "gdi32" (ByVal hdc As Long) As Long
Private Declare Function SelectObject _
Lib "gdi32" (ByVal hdc As Long, _
ByVal hObject As Long) As Long
Private Declare Function DeleteObject Lib "gdi32" (ByVal hObject As Long) As Long
Private Declare Function GetObject _
Lib "gdi32" _
Alias "GetObjectA" (ByVal hObject As Long, _
ByVal nCount As Long, _
lpObject As Any) As Long
Private Declare Function SetDIBits _
Lib "gdi32" (ByVal hdc As Long, _
ByVal hBitmap As Long, _
ByVal nStartScan As Long, _
ByVal nNumScans As Long, _
lpBits As Any, _
lpBI As BITMAPINFO, _
ByVal wUsage As Long) As Long
Private Declare Sub CopyMemory _
Lib "kernel32" _
Alias "RtlMoveMemory" (pDst As Any, _
pSrc As Any, _
ByVal ByteLen As Long)
Public Const SRCCOPY = &HCC0020
Private Type BITMAP
bmType As Long
bmWidth As Long
bmHeight As Long
bmWidthBytes As Long
bmPlanes As Integer
bmBitsPixel As Integer
bmBits As Long
End Type
Type BITMAPINFOHEADER '40 bytes
biSize As Long 'BITMAPINFOHEADER
biWidth As Long
biHeight As Long
biPlanes As Integer '
biBitCount As Integer '
biCompression As Long '
biSizeImage As Long '
biXPelsPerMeter As Long '
biYPelsPerMeter As Long '
biClrUsed As Long '
biClrImportant As Long '
End Type
Type RGBQUAD '
rgbBlue As Byte '
rgbGreen As Byte '
rgbRed As Byte '
rgbReserved As Byte '
End Type
Type BITMAPINFO
bmiHeader As BITMAPINFOHEADER
bmiColors As RGBQUAD
End Type
Private Const BI_RGB = 0&
Private Const DIB_RGB_COLORS = 0&
Private Const LR_LOADFROMFILE = &H10
Private Const IMAGE_BITMAP = 0&
Private PicInfo As BITMAP '
Private DIBInfo As BITMAPINFO 'Device Ind. Bitmap info structure
Public iDATA() As Byte
Public bDATA() As Byte '
Public Sub ReadPic(ByVal pic As Long, data() As Byte) '
Dim hdcNew As Long
Dim oldhand As Long
Dim ret As Long
Dim BytesPerScanLine As Long '
Dim PadBytesPerScanLine As Long
Call GetObject(pic, Len(PicInfo), PicInfo) '
hdcNew = CreateCompatibleDC(0&) '
oldhand = SelectObject(hdcNew, pic)
With DIBInfo.bmiHeader
.biSize = 40 '
.biWidth = PicInfo.bmWidth
.biHeight = -PicInfo.bmHeight '
.biPlanes = 1
.biBitCount = 32 '
.biCompression = BI_RGB '
BytesPerScanLine = ((((.biWidth * .biBitCount) + 31) \ 32) * 4) '
PadBytesPerScanLine = BytesPerScanLine - (((.biWidth * .biBitCount) + 7) \ 8)
.biSizeImage = BytesPerScanLine * Abs(.biHeight)
End With
ReDim data(1 To 4, 1 To PicInfo.bmWidth, 1 To PicInfo.bmHeight) As Byte
ret = GetDIBits(hdcNew, pic, 0, PicInfo.bmHeight, data(1, 1, 1), DIBInfo, DIB_RGB_COLORS)
SelectObject hdcNew, oldhand
DeleteDC hdcNew
End Sub
Public Sub ShowPic(ByVal pic As Long, data() As Byte) '
Dim hdcNew As Long
Dim oldhand As Long
Dim ret As Long
Dim BytesPerScanLine As Long '
Dim PadBytesPerScanLine As Long
Call GetObject(pic, Len(PicInfo), PicInfo) '
hdcNew = CreateCompatibleDC(0&) '
oldhand = SelectObject(hdcNew, pic)
With DIBInfo.bmiHeader
.biSize = 40 'bmp3.0
.biWidth = PicInfo.bmWidth
.biHeight = -PicInfo.bmHeight '
.biPlanes = 1
.biBitCount = 32 '
.biCompression = BI_RGB '
BytesPerScanLine = ((((.biWidth * .biBitCount) + 31) \ 32) * 4) '
PadBytesPerScanLine = BytesPerScanLine - (((.biWidth * .biBitCount) + 7) \ 8)
' .biSizeImage = BytesPerScanLine * Abs(.biHeight)
End With
ret = SetDIBits(hdcNew, pic, 0, PicInfo.bmHeight, data(1, 1, 1), DIBInfo, DIB_RGB_COLORS)
SelectObject hdcNew, oldhand
DeleteDC hdcNew
Form1.Picture3.Refresh
End Sub
Public Sub TemplateMatchDib(Lwidth As Long, Lheight As Long, Temph As Long, Tempw As Long)
Dim i As Integer, j As Integer, m As Integer, N As Integer
Dim dsigmast As Double, dsigmas As Double, dsigmat As Double, r As Double, maxr As Double
Dim LmaxWidth As Long, LmaxHeight As Long
Dim MidMatrix() As Long
ReDim MidMatrix(3, Lwidth, Lheight) As Long
On Error Resume Next
For i = 1 To Lwidth
For j = 1 To Lheight
For m = 1 To 3 '
MidMatrix(m, i, j) = 0 '
Next m
Next j
Next i
dsigmat = 0
For N = 1 To Temph - 1
For m = 1 To Tempw - 1
dsigmat = dsigmat + Val(bDATA(1, m, N)) ^ 2
Next m
Next N
maxr = 0
For j = 1 To Lheight - Temph + 1 Step 1 '
For i = 1 To Lwidth - Tempw + 1 Step 1 '
dsigmast = 0
dsigmas = 0 '
For N = 1 To Temph
For m = 1 To Tempw
dsigmas = dsigmas + (Val(iDATA(1, i + m - 1, j + N - 1)) + Val(iDATA(2, i + m - 1, j + N - 1)) + Val(iDATA(3, i + m - 1, j + N - 1)) / 3) ^ 2
dsigmast = dsigmast + (Val(iDATA(1, i + m - 1, j + N - 1)) + Val(iDATA(2, i + m - 1, j + N - 1)) + Val(iDATA(3, i + m - 1, j + N - 1)) / 3) * (Val(bDATA(1, m, N)) + Val(bDATA(1, m, N)) + Val(bDATA(1, m, N))) / 3
DoEvents
Next m
Next N
r = dsigmast / (FASTsqr(dsigmas) * Sqr(dsigmat)) '
If r > maxr Then '
maxr = r
LmaxWidth = i
LmaxHeight = j
Debug.Print r
If r > 1 Then ' 1 may or oher.
Form1.Picture1.Line (LmaxWidth, LmaxHeight)-(LmaxWidth + Tempw, LmaxHeight + Temph), vbRed, B
GoTo h:
End If
End If
DoEvents
Next i
DoEvents
Next j
h:
For N = 1 To Temph '
For m = 1 To Tempw
MidMatrix(1, m + LmaxWidth, N + LmaxHeight) = bDATA(1, m, N)
MidMatrix(2, m + LmaxWidth, N + LmaxHeight) = bDATA(2, m, N)
MidMatrix(3, m + LmaxWidth, N + LmaxHeight) = bDATA(3, m, N)
Next m
Next N
For i = 1 To Lwidth
For j = 1 To Lheight
For m = 1 To 3
iDATA(m, i, j) = MidMatrix(m, i, j)
Next m
Next j
Next i
End Sub
Public Function FASTsqr(N As Double) As Double
Dim X As Double
Dim ox As Double
If N Then
X = N * 0.5
Do
ox = X
X = (X + (N / X)) * 0.5
Loop While ox <> X
FASTsqr = X
End If
End Function
Last edited by xxdoc123; Dec 4th, 2021 at 11:53 PM.
Private Sub Command1_Click()
Picture1.ScaleMode = 3
Picture1.AutoRedraw = True
Picture1.AutoSize = True
Picture2.ScaleMode = 3
Picture2.AutoRedraw = True
Picture2.AutoSize = True
Picture3.ScaleMode = 3
Picture3.Width = Picture1.Width
Picture3.Height = Picture1.Height
Picture3.AutoRedraw = True
Picture3.AutoSize = True
If Picture2.ScaleWidth > Picture1.ScaleWidth Or Picture2.ScaleHeight > Picture1.ScaleHeight Then
MsgBox "The image size does not meet the requirements"
Exit Sub
End If
Command1.Enabled = False
Call ReadPic(Picture1.Image, iDATA())
Call ReadPic(Picture2.Image, bDATA())
' Call TemplateMatchDib(Picture1.ScaleWidth, Picture1.ScaleHeight, Picture2.ScaleHeight, Picture2.ScaleWidth)
Call FOUND(iDATA(), bDATA())
Command1.Enabled = True
Call ShowPic(Picture3.Image, iDATA)
End Sub
Code:
Public Sub ReadPic(ByVal PIC As Long, Data() As Byte) '
Dim hdcNew As Long
Dim oldhand As Long
Dim ret As Long
Dim BytesPerScanLine As Long '
Dim PadBytesPerScanLine As Long
Call GetObject(PIC, Len(PicInfo), PicInfo) '
hdcNew = CreateCompatibleDC(0&) '
oldhand = SelectObject(hdcNew, PIC)
With DIBInfo.bmiHeader
.biSize = 40 '
.biWidth = PicInfo.bmWidth
.biHeight = -PicInfo.bmHeight '
.biPlanes = 1
.biBitCount = 32 '
.biCompression = BI_RGB '
BytesPerScanLine = ((((.biWidth * .biBitCount) + 31) \ 32) * 4) '
PadBytesPerScanLine = BytesPerScanLine - (((.biWidth * .biBitCount) + 7) \ 8)
.biSizeImage = BytesPerScanLine * Abs(.biHeight)
End With
'ReDim data(1 To 4, 1 To PicInfo.bmWidth, 1 To PicInfo.bmHeight) As Byte
ReDim Data(0 To 3, 0 To PicInfo.bmWidth - 1, 0 To PicInfo.bmHeight - 1) As Byte
' ret = GetDIBits(hdcNew, pic, 0, PicInfo.bmHeight, data(1, 1, 1), DIBInfo, DIB_RGB_COLORS)
ret = GetDIBits(hdcNew, PIC, 0, PicInfo.bmHeight, Data(0, 0, 0), DIBInfo, DIB_RGB_COLORS)
SelectObject hdcNew, oldhand
DeleteDC hdcNew
End Sub
Public Sub ShowPic(ByVal PIC As Long, Data() As Byte) '
Dim hdcNew As Long
Dim oldhand As Long
Dim ret As Long
Dim BytesPerScanLine As Long '
Dim PadBytesPerScanLine As Long
Call GetObject(PIC, Len(PicInfo), PicInfo) '
hdcNew = CreateCompatibleDC(0&) '
oldhand = SelectObject(hdcNew, PIC)
With DIBInfo.bmiHeader
.biSize = 40 'bmp3.0
.biWidth = PicInfo.bmWidth
.biHeight = -PicInfo.bmHeight '
.biPlanes = 1
.biBitCount = 32 '
.biCompression = BI_RGB '
BytesPerScanLine = ((((.biWidth * .biBitCount) + 31) \ 32) * 4) '
PadBytesPerScanLine = BytesPerScanLine - (((.biWidth * .biBitCount) + 7) \ 8)
.biSizeImage = BytesPerScanLine * Abs(.biHeight)
End With
'ret = SetDIBits(hdcNew, pic, 0, PicInfo.bmHeight, data(1, 1, 1), DIBInfo, DIB_RGB_COLORS)
ret = SetDIBits(hdcNew, PIC, 0, PicInfo.bmHeight, Data(0, 0, 0), DIBInfo, DIB_RGB_COLORS)
SelectObject hdcNew, oldhand
DeleteDC hdcNew
End Sub
Code:
Public Sub FOUND(Source() As Byte, toFind() As Byte)
' COMPILE with all optimization ON. [but alias?]
Dim X&, Y&
Dim W&, H&
Dim X2&, Y2&
Dim W2&, H2&
Dim pDR As Double 'Pixel Delta
Dim pDG As Double
Dim pDB As Double
Dim tDR As Double 'Total Pixel Delta
Dim tDG As Double
Dim tDB As Double
Dim Result() As Double
Dim MinR As Double
Dim MaxR As Double
W = UBound(Source, 2)
H = UBound(Source, 3)
W2 = UBound(toFind, 2)
H2 = UBound(toFind, 3)
ReDim Result(W, H)
For X = 0 To W - W2 Step 1 'To Speed up You can play with Step
For Y = 0 To H - H2 Step 1
tDR = 0
tDG = 0
tDB = 0
For X2 = 0 To W2 Step 4 'To Speed up You can play with Step
For Y2 = 0 To H2 Step 4
pDR = Source(2, X + X2, Y + Y2) * 1 - toFind(2, X2, Y2) * 1
pDG = Source(1, X + X2, Y + Y2) * 1 - toFind(1, X2, Y2) * 1
pDB = Source(0, X + X2, Y + Y2) * 1 - toFind(0, X2, Y2) * 1
tDR = tDR + pDR * pDR
tDG = tDG + pDG * pDG
tDB = tDB + pDB * pDB
Next
Next
Result(X, Y) = tDR + tDG + tDB
If Result(X, Y) > MaxR Then MaxR = Result(X, Y)
Next
Next
MaxR = 255 / MaxR
MinR = 1E+32
For X = 0 To W - W2
For Y = 0 To H - H2
If Result(X, Y) Then
If (Result(X, Y) < MinR) Then
MinR = Result(X, Y)
X2 = X
Y2 = Y
End If
End If
Source(2, X, Y) = Result(X, Y) * MaxR
Source(1, X, Y) = Result(X, Y) * MaxR
Source(0, X, Y) = Result(X, Y) * MaxR
Next
Next
For X = 0 To W2
For Y = 0 To H2
Source(2, X2 + X, Y2 + Y) = toFind(2, X, Y)
Source(1, X2 + X, Y2 + Y) = toFind(1, X, Y)
Source(0, X2 + X, Y2 + Y) = toFind(0, X, Y)
Next
Next
End Sub
If you know in which sector of the screen the sub image that serves as the marker will appear, you could test for a couple of pixels to see if the colors match. In your example it would be easy, since you don't have red or magenta in the rest of the image.
Private Sub Command1_Click()
Picture1.ScaleMode = 3
Picture1.AutoRedraw = True
Picture1.AutoSize = True
Picture2.ScaleMode = 3
Picture2.AutoRedraw = True
Picture2.AutoSize = True
Picture3.ScaleMode = 3
Picture3.Width = Picture1.Width
Picture3.Height = Picture1.Height
Picture3.AutoRedraw = True
Picture3.AutoSize = True
If Picture2.ScaleWidth > Picture1.ScaleWidth Or Picture2.ScaleHeight > Picture1.ScaleHeight Then
MsgBox "The image size does not meet the requirements"
Exit Sub
End If
Command1.Enabled = False
Call ReadPic(Picture1.Image, iDATA())
Call ReadPic(Picture2.Image, bDATA())
' Call TemplateMatchDib(Picture1.ScaleWidth, Picture1.ScaleHeight, Picture2.ScaleHeight, Picture2.ScaleWidth)
Call FOUND(iDATA(), bDATA())
Command1.Enabled = True
Call ShowPic(Picture3.Image, iDATA)
End Sub
Code:
Public Sub ReadPic(ByVal PIC As Long, Data() As Byte) '
Dim hdcNew As Long
Dim oldhand As Long
Dim ret As Long
Dim BytesPerScanLine As Long '
Dim PadBytesPerScanLine As Long
Call GetObject(PIC, Len(PicInfo), PicInfo) '
hdcNew = CreateCompatibleDC(0&) '
oldhand = SelectObject(hdcNew, PIC)
With DIBInfo.bmiHeader
.biSize = 40 '
.biWidth = PicInfo.bmWidth
.biHeight = -PicInfo.bmHeight '
.biPlanes = 1
.biBitCount = 32 '
.biCompression = BI_RGB '
BytesPerScanLine = ((((.biWidth * .biBitCount) + 31) \ 32) * 4) '
PadBytesPerScanLine = BytesPerScanLine - (((.biWidth * .biBitCount) + 7) \ 8)
.biSizeImage = BytesPerScanLine * Abs(.biHeight)
End With
'ReDim data(1 To 4, 1 To PicInfo.bmWidth, 1 To PicInfo.bmHeight) As Byte
ReDim Data(0 To 3, 0 To PicInfo.bmWidth - 1, 0 To PicInfo.bmHeight - 1) As Byte
' ret = GetDIBits(hdcNew, pic, 0, PicInfo.bmHeight, data(1, 1, 1), DIBInfo, DIB_RGB_COLORS)
ret = GetDIBits(hdcNew, PIC, 0, PicInfo.bmHeight, Data(0, 0, 0), DIBInfo, DIB_RGB_COLORS)
SelectObject hdcNew, oldhand
DeleteDC hdcNew
End Sub
Public Sub ShowPic(ByVal PIC As Long, Data() As Byte) '
Dim hdcNew As Long
Dim oldhand As Long
Dim ret As Long
Dim BytesPerScanLine As Long '
Dim PadBytesPerScanLine As Long
Call GetObject(PIC, Len(PicInfo), PicInfo) '
hdcNew = CreateCompatibleDC(0&) '
oldhand = SelectObject(hdcNew, PIC)
With DIBInfo.bmiHeader
.biSize = 40 'bmp3.0
.biWidth = PicInfo.bmWidth
.biHeight = -PicInfo.bmHeight '
.biPlanes = 1
.biBitCount = 32 '
.biCompression = BI_RGB '
BytesPerScanLine = ((((.biWidth * .biBitCount) + 31) \ 32) * 4) '
PadBytesPerScanLine = BytesPerScanLine - (((.biWidth * .biBitCount) + 7) \ 8)
.biSizeImage = BytesPerScanLine * Abs(.biHeight)
End With
'ret = SetDIBits(hdcNew, pic, 0, PicInfo.bmHeight, data(1, 1, 1), DIBInfo, DIB_RGB_COLORS)
ret = SetDIBits(hdcNew, PIC, 0, PicInfo.bmHeight, Data(0, 0, 0), DIBInfo, DIB_RGB_COLORS)
SelectObject hdcNew, oldhand
DeleteDC hdcNew
End Sub
Code:
Public Sub FOUND(Source() As Byte, toFind() As Byte)
' COMPILE with all optimization ON. [but alias?]
Dim X&, Y&
Dim W&, H&
Dim X2&, Y2&
Dim W2&, H2&
Dim pDR As Double 'Pixel Delta
Dim pDG As Double
Dim pDB As Double
Dim tDR As Double 'Total Pixel Delta
Dim tDG As Double
Dim tDB As Double
Dim Result() As Double
Dim MinR As Double
Dim MaxR As Double
W = UBound(Source, 2)
H = UBound(Source, 3)
W2 = UBound(toFind, 2)
H2 = UBound(toFind, 3)
ReDim Result(W, H)
For X = 0 To W - W2 Step 1 'To Speed up You can play with Step
For Y = 0 To H - H2 Step 1
tDR = 0
tDG = 0
tDB = 0
For X2 = 0 To W2 Step 4 'To Speed up You can play with Step
For Y2 = 0 To H2 Step 4
pDR = Source(2, X + X2, Y + Y2) * 1 - toFind(2, X2, Y2) * 1
pDG = Source(1, X + X2, Y + Y2) * 1 - toFind(1, X2, Y2) * 1
pDB = Source(0, X + X2, Y + Y2) * 1 - toFind(0, X2, Y2) * 1
tDR = tDR + pDR * pDR
tDG = tDG + pDG * pDG
tDB = tDB + pDB * pDB
Next
Next
Result(X, Y) = tDR + tDG + tDB
If Result(X, Y) > MaxR Then MaxR = Result(X, Y)
Next
Next
MaxR = 255 / MaxR
MinR = 1E+32
For X = 0 To W - W2
For Y = 0 To H - H2
If Result(X, Y) Then
If (Result(X, Y) < MinR) Then
MinR = Result(X, Y)
X2 = X
Y2 = Y
End If
End If
Source(2, X, Y) = Result(X, Y) * MaxR
Source(1, X, Y) = Result(X, Y) * MaxR
Source(0, X, Y) = Result(X, Y) * MaxR
Next
Next
For X = 0 To W2
For Y = 0 To H2
Source(2, X2 + X, Y2 + Y) = toFind(2, X, Y)
Source(1, X2 + X, Y2 + Y) = toFind(1, X, Y)
Source(0, X2 + X, Y2 + Y) = toFind(0, X, Y)
Next
Next
End Sub
Your code looks more standardized.However, this method of overall comparison of chromatic aberrations still has certain flaws. Visually completely different images will appear the same difference.
Code:
For X = 0 To W - W2 Step 1 'To Speed up You can play with Step
For Y = 0 To H - H2 Step 1
tDR = 0
tDG = 0
tDB = 0
For X2 = 0 To W2 Step 4 'To Speed up You can play with Step
For Y2 = 0 To H2 Step 4
pDR = Source(2, X + X2, Y + Y2) * 1 - toFind(2, X2, Y2) * 1
pDG = Source(1, X + X2, Y + Y2) * 1 - toFind(1, X2, Y2) * 1
pDB = Source(0, X + X2, Y + Y2) * 1 - toFind(0, X2, Y2) * 1
if abs(pDR )<50 and abs(pDG )<50 and abs(pDB )<50 then ' 50 is Color difference'
sameNum =sameNum+1
end if
Next
Next
if sameNum >W2*H2* 0.9 then '0.9 is Similarity ratio
msgbox "find"
sameNum =0
end if
Next
Next
If you know in which sector of the screen the sub image that serves as the marker will appear, you could test for a couple of pixels to see if the colors match. In your example it would be easy, since you don't have red or magenta in the rest of the image.
you are right. If the color difference between the target and the background is obvious, the color finding is indeed faster and accurate.
I don’t know if there is an algorithm for image contour comparison. In this way, through grayscale, binarization, and contour comparison, you can ignore the difference in color. It seems that this is a bit more difficult. And if the target image is not the same size as the template image
Inspired by your question I started to work in am image comparison tool (that BTW it is already more or less working).
But reading again your OP and being downloaded your sample images, I realize that an Image comparison tool like the one I'm doing won't serve for what you want.
You need to recognize a small image into a larger one, and that's not what my current project does.
For what you need, if you know the place where the smaller image will be, it could be quite easy to check whether it is there or not.
As I said, I would just check for a couple of pixels.
Inspired by your question I started to work in am image comparison tool (that BTW it is already more or less working).
But reading again your OP and being downloaded your sample images, I realize that an Image comparison tool like the one I'm doing won't serve for what you want.
You need to recognize a small image into a larger one, and that's not what my current project does.
For what you need, if you know the place where the smaller image will be, it could be quite easy to check whether it is there or not.
As I said, I would just check for a couple of pixels.
... However, this method of overall comparison of chromatic aberrations still has certain flaws. Visually completely different images will appear the same difference.
I'm not so convinced about that.
However, I was inspired to try to calculate the result based on the edges.
I tried using just their magnitudes but it didn't seem to work well. (perhaps for similar reasons to what you were saying).
So I added the orientation of the edges.
This seems to work fine.
I haven't done much testing though.
Code:
Public Const PI As Double = 3.14159265358979
Public Const PIh As Double = 1.5707963267949
Public Const PI2 As Double = 6.28318530717959
Private Function Atan2(X As Double, Y As Double) As Double
If X Then
Atan2 = -PI + Atn(Y / X) - (X > 0#) * PI
Else
Atan2 = -PIh - (Y > 0#) * PI
End If
End Function
Private Function AngleDIFF(A1 As Double, A2 As Double) As Double
AngleDIFF = A1 - A2
While AngleDIFF < -PI
AngleDIFF = AngleDIFF + PI2
Wend
While AngleDIFF > PI
AngleDIFF = AngleDIFF - PI2
Wend
End Function
Public Sub FOUND2(Source() As Byte, toFind() As Byte)
Dim X&, Y&
Dim W&, H&
Dim X2&, Y2&
Dim W2&, H2&
Dim pDR As Double
Dim pDG As Double
Dim pDB As Double
Dim tDR As Double
Dim tDG As Double
Dim tDB As Double
Dim Result() As Double
Dim EdgeAng1() As Double
Dim EdgeAng2() As Double
Dim MAG2() As Double
Dim MAG1() As Double
Dim DX#, DY#
Dim MinR As Double
Dim MaxR As Double
W = UBound(Source, 2)
H = UBound(Source, 3)
W2 = UBound(toFind, 2)
H2 = UBound(toFind, 3)
ReDim Result(W, H)
ReDim EdgeAng1(W, H)
ReDim EdgeAng2(W2, H2)
ReDim MAG1(W, H)
ReDim MAG2(W2, H2) 'EDGE magnitude
' Convert to Edge Angle --------------------------
For Y = 1 To H - 1
For X = 1 To W - 1
pDR = Source(2, X - 1, Y) * 1 - Source(2, X + 1, Y) * 1
pDG = Source(1, X - 1, Y) * 1 - Source(1, X + 1, Y) * 1
pDB = Source(0, X - 1, Y) * 1 - Source(0, X + 1, Y) * 1
DX = (pDR) + (pDG) + (pDB)
pDR = Source(2, X, Y - 1) * 1 - Source(2, X, Y + 1) * 1
pDG = Source(1, X, Y - 1) * 1 - Source(1, X, Y + 1) * 1
pDB = Source(0, X, Y - 1) * 1 - Source(0, X, Y + 1) * 1
DY = (pDR) + (pDG) + (pDB)
EdgeAng1(X, Y) = Atan2(DX, DY)
MAG1(X, Y) = Sqr(DX * DX + DY * DY)
Next
Next
For Y = 1 To H2 - 1
For X = 1 To W2 - 1
pDR = toFind(2, X - 1, Y) * 1 - toFind(2, X + 1, Y) * 1
pDG = toFind(1, X - 1, Y) * 1 - toFind(1, X + 1, Y) * 1
pDB = toFind(0, X - 1, Y) * 1 - toFind(0, X + 1, Y) * 1
DX = (pDR) + (pDG) + (pDB)
pDR = toFind(2, X, Y - 1) * 1 - toFind(2, X, Y + 1) * 1
pDG = toFind(1, X, Y - 1) * 1 - toFind(1, X, Y + 1) * 1
pDB = toFind(0, X, Y - 1) * 1 - toFind(0, X, Y + 1) * 1
DY = (pDR) + (pDG) + (pDB)
EdgeAng2(X, Y) = Atan2(DX, DY)
MAG2(X, Y) = Sqr(DX * DX + DY * DY)
Next
Next
'---------------------------------
For X = 0 To W - W2
For Y = 0 To H - H2
tDR = 0
For X2 = 0 To W2 Step 4
For Y2 = 0 To H2 Step 4
pDR = AngleDIFF(EdgeAng1(X + X2, Y + Y2), EdgeAng2(X2, Y2))
tDR = tDR + pDR * pDR * MAG2(X2, Y2)
Next
Next
Result(X, Y) = tDR
If Result(X, Y) > MaxR Then MaxR = Result(X, Y)
Next
Next
MaxR = 255 / MaxR
MinR = 1E+32
For X = 0 To W - W2
For Y = 0 To H - H2
If Result(X, Y) Then
If (Result(X, Y) < MinR) Then
MinR = Result(X, Y)
X2 = X
Y2 = Y
End If
End If
Source(2, X, Y) = Result(X, Y) * MaxR
Source(1, X, Y) = Result(X, Y) * MaxR
Source(0, X, Y) = Result(X, Y) * MaxR
Next
Next
For X = 0 To W2
For Y = 0 To H2
Source(2, X2 + X, Y2 + Y) = toFind(2, X, Y)
Source(1, X2 + X, Y2 + Y) = toFind(1, X, Y)
Source(0, X2 + X, Y2 + Y) = toFind(0, X, Y)
Next
Next
End Sub
EDIT
Code:
pDR = toFind(2, X, Y - 1) * 1 - toFind(2, X, Y + 1) * 1
pDG = toFind(1, X, Y - 1) * 1 - toFind(1, X, Y + 1) * 1
pDB = toFind(0, X, Y - 1) * 1 - toFind(0, X, Y + 1) * 1
I'm not so convinced about that.
However, I was inspired to try to calculate the result based on the edges.
I tried using just their magnitudes but it didn't seem to work well. (perhaps for similar reasons to what you were saying).
So I added the orientation of the edges.
This seems to work fine.
I haven't done much testing though.
Code:
Public Const PI As Double = 3.14159265358979
Public Const PIh As Double = 1.5707963267949
Public Const PI2 As Double = 6.28318530717959
Private Function Atan2(X As Double, Y As Double) As Double
If X Then
Atan2 = -PI + Atn(Y / X) - (X > 0#) * PI
Else
Atan2 = -PIh - (Y > 0#) * PI
End If
End Function
Private Function AngleDIFF(A1 As Double, A2 As Double) As Double
AngleDIFF = A1 - A2
While AngleDIFF < -PI
AngleDIFF = AngleDIFF + PI2
Wend
While AngleDIFF > PI
AngleDIFF = AngleDIFF - PI2
Wend
End Function
Public Sub FOUND2(Source() As Byte, toFind() As Byte)
Dim X&, Y&
Dim W&, H&
Dim X2&, Y2&
Dim W2&, H2&
Dim pDR As Double
Dim pDG As Double
Dim pDB As Double
Dim tDR As Double
Dim tDG As Double
Dim tDB As Double
Dim Result() As Double
Dim EdgeAng1() As Double
Dim EdgeAng2() As Double
Dim MAG2() As Double
Dim MAG1() As Double
Dim DX#, DY#
Dim MinR As Double
Dim MaxR As Double
W = UBound(Source, 2)
H = UBound(Source, 3)
W2 = UBound(toFind, 2)
H2 = UBound(toFind, 3)
ReDim Result(W, H)
ReDim EdgeAng1(W, H)
ReDim EdgeAng2(W2, H2)
ReDim MAG1(W, H)
ReDim MAG2(W2, H2) 'EDGE magnitude
' Convert to Edge Angle --------------------------
For Y = 1 To H - 1
For X = 1 To W - 1
pDR = Source(2, X - 1, Y) * 1 - Source(2, X + 1, Y) * 1
pDG = Source(1, X - 1, Y) * 1 - Source(1, X + 1, Y) * 1
pDB = Source(0, X - 1, Y) * 1 - Source(0, X + 1, Y) * 1
DX = (pDR) + (pDG) + (pDB)
pDR = Source(2, X, Y - 1) * 1 - Source(2, X, Y + 1) * 1
pDG = Source(1, X, Y - 1) * 1 - Source(1, X, Y + 1) * 1
pDB = Source(0, X, Y - 1) * 1 - Source(0, X, Y + 1) * 1
DY = (pDR) + (pDG) + (pDB)
EdgeAng1(X, Y) = Atan2(DX, DY)
MAG1(X, Y) = Sqr(DX * DX + DY * DY)
Next
Next
For Y = 1 To H2 - 1
For X = 1 To W2 - 1
pDR = toFind(2, X - 1, Y) * 1 - toFind(2, X + 1, Y) * 1
pDG = toFind(1, X - 1, Y) * 1 - toFind(1, X + 1, Y) * 1
pDB = toFind(0, X - 1, Y) * 1 - toFind(0, X + 1, Y) * 1
DX = (pDR) + (pDG) + (pDB)
pDR = toFind(2, X, Y - 1) * 1 - toFind(2, X, Y + 1) * 1
pDG = toFind(1, X, Y - 1) * 1 - toFind(1, X, Y + 1) * 1
pDB = toFind(0, X, Y - 1) * 1 - toFind(0, X, Y + 1) * 1
DY = (pDR) + (pDG) + (pDB)
EdgeAng2(X, Y) = Atan2(DX, DY)
MAG2(X, Y) = Sqr(DX * DX + DY * DY)
Next
Next
'---------------------------------
For X = 0 To W - W2
For Y = 0 To H - H2
tDR = 0
For X2 = 0 To W2 Step 4
For Y2 = 0 To H2 Step 4
pDR = AngleDIFF(EdgeAng1(X + X2, Y + Y2), EdgeAng2(X2, Y2))
tDR = tDR + pDR * pDR * MAG2(X2, Y2)
Next
Next
Result(X, Y) = tDR
If Result(X, Y) > MaxR Then MaxR = Result(X, Y)
Next
Next
MaxR = 255 / MaxR
MinR = 1E+32
For X = 0 To W - W2
For Y = 0 To H - H2
If Result(X, Y) Then
If (Result(X, Y) < MinR) Then
MinR = Result(X, Y)
X2 = X
Y2 = Y
End If
End If
Source(2, X, Y) = Result(X, Y) * MaxR
Source(1, X, Y) = Result(X, Y) * MaxR
Source(0, X, Y) = Result(X, Y) * MaxR
Next
Next
For X = 0 To W2
For Y = 0 To H2
Source(2, X2 + X, Y2 + Y) = toFind(2, X, Y)
Source(1, X2 + X, Y2 + Y) = toFind(1, X, Y)
Source(0, X2 + X, Y2 + Y) = toFind(0, X, Y)
Next
Next
End Sub
EDIT
Code:
pDR = toFind(2, X, Y - 1) * 1 - toFind(2, X, Y + 1) * 1
pDG = toFind(1, X, Y - 1) * 1 - toFind(1, X, Y + 1) * 1
pDB = toFind(0, X, Y - 1) * 1 - toFind(0, X, Y + 1) * 1
The effect is good.i test binarization, grayscale and other matching can also be well supported
Windows message simulation, common keyboard message and mouse message simulation.
Support common screenshots, GDI, DX (including d3d9, D3D10, d3d11), OpenGL screenshots, and minimize screenshots of common simulators (lightning, night God)
Find color and map, support color deviation and fuzzy recognition
Character recognition (OCR) supports 255 x 255 large dot matrix at most, color deviation, fuzzy recognition, system font library and desert font library
The plug-in has 32-bit and 64 bit versions and supports 32 / 64 bit binding
The project is completely open source, no back door, no virus, and can be used safely
I don’t know if there is an algorithm for image contour comparison.
If you can convert the image to b/w and know its position on screen then you can use histogram to detect the image. I wrote a simple image recognition program using this technique working well to read playing cards captured by a camera.
I sliced the area (not necessarily to scale) then saved number of black pixels in each rectangle. Then the black pixels in the same area of the captured image is counted, difference from the reference is squared and errors are added up. If it is below some value for a defined shape then that shape is decided.
Attached picture is for 45 x 45 pixels apple, very randomly sliced just to give an idea. You can decide how to slice to get a better estimation. You can expect quite a difference between apple and pear .
If you can convert the image to b/w and know its position on screen then you can use histogram to detect the image. I wrote a simple image recognition program using this technique working well to read playing cards captured by a camera.
I sliced the area (not necessarily to scale) then saved number of black pixels in each rectangle. Then the black pixels in the same area of the captured image is counted, difference from the reference is squared and errors are added up. If it is below some value for a defined shape then that shape is decided.
Attached picture is for 45 x 45 pixels apple, very randomly sliced just to give an idea. You can decide how to slice to get a better estimation. You can expect quite a difference between apple and pear .
I searched a lot of algorithms. But many are not in vb language.
I didn't write a vb program for it, wrote only assembler for microcontroller. But I will write the algorithm with two sample images, the club and heart of playing cards, resized to 48x64 pixels (heart is stretced so aspect ratio is not as original but it shouldn't be required for your project).
The images are divided into 4 rows and 3 columns of 16x16 pixels each (background colored for indication). So each image has 12 histogram values, starting from upper left to right :
Club : 21, 244, 17, 88, 208, 88, 250, 228, 250, 67, 76, 67 black pixels (not precisely counted).
Heart : 224, 181, 224, 229, 256, 229, 122, 256, 122, 6, 172, 6 black pixels.
You can read the pixel by Point instruction to count blacks in desired region (or any desired color if working with colors).
Then calculate sum of squares of differences for each area. An example with a distorted club shape is also given.
Distorted club : 40, 233, 7, 79, 198, 60, 256, 236, 251, 99, 105, 30 black pixels.
Divergence from reference club : (21-40)^2+(244-233)^2+... = 4882
Divergence from reference heart : (224-40)^2+(181-233)^2+... = 248761
You will see a lot of difference, enough to decide the shape. I hope it gived an idea how you can detect an image.
I didn't write a vb program for it, wrote only assembler for microcontroller. But I will write the algorithm with two sample images, the club and heart of playing cards, resized to 48x64 pixels (heart is stretced so aspect ratio is not as original but it shouldn't be required for your project).
The images are divided into 4 rows and 3 columns of 16x16 pixels each (background colored for indication). So each image has 12 histogram values, starting from upper left to right :
Club : 21, 244, 17, 88, 208, 88, 250, 228, 250, 67, 76, 67 black pixels (not precisely counted).
Heart : 224, 181, 224, 229, 256, 229, 122, 256, 122, 6, 172, 6 black pixels.
You can read the pixel by Point instruction to count blacks in desired region (or any desired color if working with colors).
Then calculate sum of squares of differences for each area. An example with a distorted club shape is also given.
Distorted club : 40, 233, 7, 79, 198, 60, 256, 236, 251, 99, 105, 30 black pixels.
Divergence from reference club : (21-40)^2+(244-233)^2+... = 4882
Divergence from reference heart : (224-40)^2+(181-233)^2+... = 248761
You will see a lot of difference, enough to decide the shape. I hope it gived an idea how you can detect an image.
Yes, it is so, if only global comparison will be done no need to calculate squareroot. Since I was writing for a limited memory (and speed) microcontroller I didn't use squareroot.
Yes, it is so, if only global comparison will be done no need to calculate squareroot. Since I was writing for a limited memory (and speed) microcontroller I didn't use squareroot.
Public Sub FOUND3(Source() As Byte, toFind() As Byte)
' COMPILE with all optimization ON. [but alias?]
Dim X&, Y&
Dim W&, H&
Dim X2&, Y2&
Dim W2&, H2&
Dim X3&, Y3&
Dim pDR As Double 'Pixel Delta
Dim pDG As Double
Dim pDB As Double
Dim tDR As Double 'Total Pixel Delta
Dim tDG As Double
Dim tDB As Double
Dim Result() As Double
Dim MinR As Double
Dim MaxR As Double
Dim scan0 As String
Dim scan1 As String
W = UBound(Source, 2)
H = UBound(Source, 3)
W2 = UBound(toFind, 2)
H2 = UBound(toFind, 3)
ReDim Result(W , H )
MinR = 1E+32
For X = 0 To W - W2 Step 4 'To Speed up You can play with Step
For Y = 0 To H - H2 Step 4
tDR = 0
tDG = 0
tDB = 0
For X2 = 0 To W2 - 1 Step 6 'To Speed up You can play with Step
For Y2 = 0 To H2 - 1 Step 6
pDR = Source(2, X + X2, Y + Y2)
pDG = Source(1, X + X2, Y + Y2)
pDB = Source(0, X + X2, Y + Y2)
If RGB(pDR, pDG, pDB) = 0 Then 'IF VBBLACK
scan0 = scan0 & "1"
Else
scan0 = scan0 & "0"
End If
If RGB(toFind(2, X2, Y2), toFind(1, X2, Y2), toFind(0, X2, Y2)) = 0 Then
scan1 = scan1 & "1"
Else
scan1 = scan1 & "0"
End If
Next
Next
Result(X, Y) = compareHamming(scan1, scan0)
scan1 = ""
scan0 = ""
If Result(X, Y) <= 20 Then
Form1.Picture1.Line (X, Y)-(X + W2, Y + H2), vbRed, B
End If
If Result(X, Y) > MaxR Then MaxR = Result(X, Y)
If (Result(X, Y) < MinR) Then
MinR = Result(X, Y)
X3 = X
Y3 = Y
End If
Next
Next
MaxR = 255 / MaxR
For X = 0 To W - W2
For Y = 0 To H - H2
Source(2, X, Y) = Result(X, Y) * MaxR
Source(1, X, Y) = Result(X, Y) * MaxR
Source(0, X, Y) = Result(X, Y) * MaxR
Next
Next
Debug.Print Result(X, Y)
Form1.Picture1.Line (X3, Y3)-(X3 + W2, Y3 + H2), vbGreen, B
For X = 0 To W2
For Y = 0 To H2
Source(2, X3 + X, Y3 + Y) = toFind(2, X, Y)
Source(1, X3 + X, Y3 + Y) = toFind(1, X, Y)
Source(0, X3 + X, Y3 + Y) = toFind(0, X, Y)
Next
Next
End Sub
Private Function compareHamming(inpA As String, inpB As String)
Dim HD As Long
HD = Hamming(inpA, inpB)
Select Case HD
MsgBox "Input strings should be of the same length"
Case -2
MsgBox "Input 'A' contains invalid characters; should be 0's or 1's"
Case -3
MsgBox "Input 'B' contains invalid characters; should be 0's or 1's"
Case -4
MsgBox "Both Inputs were invalid"
Case Else
compareHamming = HD
End Select
End Function
Private Function Hamming(ByVal StrA As String, ByVal StrB As String) As Long
Dim TmpBin As String
Dim i As Long
If Len(StrA) <> Len(StrB) Then
Hamming = -1
Exit Function
End If
TmpBin = Replace(StrA, "1", "")
TmpBin = Replace(TmpBin, "0", "")
If Len(TmpBin) <> 0 Then
Hamming = -2
End If
TmpBin = Replace(StrB, "1", "")
TmpBin = Replace(TmpBin, "0", "")
If Len(TmpBin) <> 0 Then
If Hamming = 0 Then 'StrA was valid
Hamming = -3
Else
Hamming = -4
End If
End If
If Hamming <> 0 Then Exit Function
For i = 1 To Len(StrA)
If Mid$(StrA, i, 1) <> Mid$(StrB, i, 1) Then
Hamming = Hamming + 1
End If
Next i
End Function
Case -1
Last edited by xxdoc123; Dec 16th, 2021 at 09:53 PM.
This algorithm can find a similar object which not needs to be exactly the same. Please see the picture below.
However, the way to use OpenCV in VB6 is not easy. So, finally, I developed the DLL in C++, then call through VB6 instead. The speed is better than VB6 coding very much, but still got some lack if the main picture is too big.
I can share my dll for you if need. But if you are still having fun with your research. That's should be fine.
*I have limited private time. So, will provide the DLL once you ask for it.
Public Function compareStr(sw1 As String, sw2 As String) As Double
Dim H As Integer, N As Long
Dim SB1 As String, SB2 As String
Debug.Print sw1
Debug.Print sw2
For H = 1 To Len(sw1)
' If H = Len(sw1) Then Stop
SB1 = Mid(sw1, H, 1)
SB2 = Mid(sw2, H, 1)
If SB1 = SB2 Then
N = N + 1 'same point
End If
Next
compareStr = N / Len(sw1)
If compareStr = 1 Then
'MsgBox "Completely similar" & Chr(10) & sw1 & Chr(10) & sw2
ElseIf compareStr > 0.9 Then
' MsgBox "very similar" & Chr(10) & sw1 & Chr(10) & sw2
ElseIf compareStr > 0.8 Then
'MsgBox "Basically similar" & Chr(10) & sw1 & Chr(10) & sw2
Else
'MsgBox "not similar" & Chr(10) & sw1 & Chr(10) & sw2
End If
End Function
This algorithm can find a similar object which not needs to be exactly the same. Please see the picture below.
However, the way to use OpenCV in VB6 is not easy. So, finally, I developed the DLL in C++, then call through VB6 instead. The speed is better than VB6 coding very much, but still got some lack if the main picture is too big.
I can share my dll for you if need. But if you are still having fun with your research. That's should be fine.
*I have limited private time. So, will provide the DLL once you ask for it.
openCV Very useful.Did not see the vb code。THANKS YOU .I have sent you a message
** If the project does not run. Please contact me anytime. Since never try to run on another computer before.
*** The project is quite big (from DLLs). So, I upload it to my OneDrive instead.
** If the project does not run. Please contact me anytime. Since never try to run on another computer before.
*** The project is quite big (from DLLs). So, I upload it to my OneDrive instead.
1. Make sure you installed the Visual C++ Redistributable 2017 from the link by previous post.
2. If can't find. Please copy all OpenCV DLLs from sub folder to project folder.
copy all OpenCV DLLs from sub folder to project folder. WORK FINE。 It's so fast. Does IFWOpenCV.dll have other functions
Sure, many of them. BTW, not much time to explain to you. Please check the OpenCV website then tell me which functions of OpenCV you are interested in. I'm not implementing all OpenCV functions since I also developed my own image processing function to serve my software.
Yes it is too complicated. What if the image you are looking for is zoomed. Seems to be inaccurate
Template1.bmp 127x125.is Zoomed to 100x60
So far, most of these functions work with controlled environments e.g. specific cameras, specific sizes of images. If you want a better functions. Please look for FLANN functions or deep learning at:-
However these functions, they said it is commercial functions that need to pay to use it. So, the normal C++ compiler as I do will skip to compile it. (Or I still have not found the way to work with it.)
BTW, look that the Python and Linux can work well with it. I never use Python. So, never try that too.
Reply With Quote
Originally Posted by reexre
.
