Skip to content

fix: pixelocr.recognize lines should be much more reliable now #545

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
Torwent wants to merge 1 commit into from
Closed

fix: pixelocr.recognize lines should be much more reliable now #545

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
57 changes: 37 additions & 20 deletions Source/simba.pixelocr.pas
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -77,6 +77,9 @@ TPixelOCR = record

const
ALPHA_NUM_SYMBOLS = ['a'..'z', 'A'..'Z', '0'..'9', '%', '&', '#', '$', '[', ']', '{', '}', '@', '!', '?'];
var
AlphaNumSymbolTableInitialized: Boolean = False;
AlphaNumSymbolTable: array [Char] of Boolean;

implementation

Expand Down Expand Up @@ -112,13 +115,21 @@ function GetGlyph(const Font: PPixelFont; const c: Char): PPixelFontGlyph; inlin
Result := nil;
end;

function ContainsAlphaNumSym(const Text: string): Boolean; inline;
function IsAlphaNumSym(const text: string): Boolean; inline;
var
I: Integer;
i, count, needed: Integer;
begin
for I := 1 to Length(Text) do
if Text[I] in ALPHA_NUM_SYMBOLS then
Exit(True);
count := 0;
needed := (Length(text) * 50 + 99) div 100;

for i := 1 to Length(text) do
if AlphaNumSymbolTable[text[i]] then
begin
Inc(count);
//if 50% of the text is ALPHA_NUM_SYMBOLS we are good.
if count >= needed then
Exit(True);
end;

Result := False;
end;
Expand Down Expand Up @@ -251,6 +262,7 @@ function TPixelOCR._RecognizeX(const Image: TSimbaImage; const Font: PPixelFont;
Result.Bounds.Y1 := $FFFFFF;
Space := 0;

BestGlyph := @Font^.Glyphs[0];
Lo := @Font^.Glyphs[0];
Hi := @Font^.Glyphs[High(Font^.Glyphs)];

Expand Down Expand Up @@ -578,42 +590,46 @@ function TPixelOCR.Recognize(Image: TSimbaImage; constref Font: TPixelFont; Boun
end;

function TPixelOCR.RecognizeLines(Image: TSimbaImage; constref Font: TPixelFont; Bounds: TBox): TStringArray;

var Temp: TPixelOCR;
function MaybeRecognize(const X, Y: Integer; const isBinary: Boolean; out Match: TPixelOCRMatch): Boolean;
var
Temp: TPixelOCR;
begin
Result := False;

// use a copy here since we change these properties
Temp := Self;
Temp.Whitelist := ALPHA_NUM_SYMBOLS;
Temp.MaxLen := 1;
Temp.MaxWalk := 0;

// Find something on a row that isn't a small character
Match := Temp._RecognizeX(Image, @Font, X, Y, isBinary);
if (Match.Hits > 0) then
begin
// OCR the row and some extra rows
Temp.Whitelist := Self.Whitelist;
Temp.MaxWalk := 0;
Temp.MaxLen := 0;

Match := Temp._RecognizeXY(Image, @Font, X, Y, Font.MaxGlyphHeight div 2, isBinary);
// Ensure that actual Text was extracted, not just a symbol mess of short or small character symbols.
if ContainsAlphaNumSym(Match.Text) then
if IsAlphaNumSym(Match.Text) then
Result := True;
end;
end;

var
isBinary: Boolean;
Match: TPixelOCRMatch;
c: Char;
begin
if (Length(Font.Glyphs) = 0) then
SimbaException('Font is empty');

if not AlphaNumSymbolTableInitialized then
begin
FillChar(AlphaNumSymbolTable, SizeOf(AlphaNumSymbolTable), False);
for c in ALPHA_NUM_SYMBOLS do
AlphaNumSymbolTable[c] := True;
AlphaNumSymbolTableInitialized := True;
end;

Temp := Self;
Temp.Whitelist := ALPHA_NUM_SYMBOLS;
Temp.MaxLen := 1;
Temp.MaxWalk := 0;
Temp.Whitelist := Self.Whitelist;
Temp.MaxWalk := 0;
Temp.MaxLen := 0;

Result := [];
Matches := [];

Expand All @@ -634,6 +650,7 @@ function TPixelOCR.RecognizeLines(Image: TSimbaImage; constref Font: TPixelFont;
// Now we can confidently skip this search line by a jump, but we dont skip fully in case of close/overlapping Text
// So we divide the texts max glyph Height by 2 and subtract that from the lower end of the found bounds.
Bounds.Y1 := Max(Bounds.Y1, Match.Bounds.Y2 - (Font.MaxGlyphHeight div 2));
Continue;
end;

Bounds.Y1 += 1;
Expand Down