Marmelator/layout-parser
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 289 Wiki Activity

RED-8642: Use LineMode from cv-analysis-service instead of table cell mode

Browse Source 
* changes for line mode instead of cell mode of cv table parsing
This commit is contained in:
maverickstuder 2024-03-07 13:31:56 +01:00
parent e99fad0c3d
commit ad1e44ca5c
4 changed files with 13 additions and 39 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/LayoutParsingPipeline.java
View File

@ -282,7 +282,7 @@ public class LayoutParsingPipeline {
} }
if (signatures.containsKey(pageNumber)) { if (signatures.containsKey(pageNumber)) {
if (classificationPage.getImages() == null || classificationPage.getImages().size() == 0) { if (classificationPage.getImages() == null || classificationPage.getImages().isEmpty()) {
classificationPage.setImages(signatures.get(pageNumber)); classificationPage.setImages(signatures.get(pageNumber));
} else { } else {
classificationPage.getImages().addAll(signatures.get(pageNumber)); classificationPage.getImages().addAll(signatures.get(pageNumber));

2
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/adapter/CvTableParsingAdapter.java
View File

@ -6,13 +6,11 @@ import java.util.Collection;
import java.util.HashMap; import java.util.HashMap;
import java.util.List; import java.util.List;
import java.util.Map; import java.util.Map;
import java.util.stream.Collectors;
import org.springframework.stereotype.Service; import org.springframework.stereotype.Service;
import com.knecon.fforesight.service.layoutparser.processor.model.table.Ruling; import com.knecon.fforesight.service.layoutparser.processor.model.table.Ruling;
import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.PageInfo; import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.PageInfo;
import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableCell;
import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableLine; import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableLine;
import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse; import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse;

2
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/table/TableData.java
View File

@ -14,7 +14,7 @@ import lombok.NoArgsConstructor;
@AllArgsConstructor @AllArgsConstructor
public class TableData { public class TableData {
private PageInfo pageInfo;; private PageInfo pageInfo;
private List<TableLine> tableLines = new ArrayList<>(); private List<TableLine> tableLines = new ArrayList<>();
} }

46
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/RulingCleaningService.java
View File

@ -35,10 +35,10 @@ public class RulingCleaningService {
Rulings verticalAndHorizontalRulingLines; Rulings verticalAndHorizontalRulingLines;
// todo 8642: set cv parsed only when no rulings exist or just always add them? // todo 8642: set cv parsed only when no rulings exist or just always add them?
if (!rulings.isEmpty()) { if (rulings.isEmpty() && parsedRulings != null) {
verticalAndHorizontalRulingLines = extractVerticalAndHorizontalRulingLines(rulings);
} else {
verticalAndHorizontalRulingLines = extractVerticalAndHorizontalRulingLines(parsedRulings); verticalAndHorizontalRulingLines = extractVerticalAndHorizontalRulingLines(parsedRulings);
} else {
verticalAndHorizontalRulingLines = extractVerticalAndHorizontalRulingLines(rulings);
} }
verticalAndHorizontalRulingLines.verticalLines.sort(X_FIRST_RULING_COMPARATOR); verticalAndHorizontalRulingLines.verticalLines.sort(X_FIRST_RULING_COMPARATOR);
@ -81,7 +81,7 @@ public class RulingCleaningService {
Rectangle rectangle2 = rectangles.get(j); Rectangle rectangle2 = rectangles.get(j);
// we can stop early when we are too far off because of x-y-sorting // we can stop early when we are too far off because of x-y-sorting
if(rectangle1.getRight() < rectangle2.getLeft() && rectangle1.getBottom() < rectangle2.getTop()) { if (rectangle1.getRight() < rectangle2.getLeft() && rectangle1.getBottom() < rectangle2.getTop()) {
break; break;
} }
@ -159,43 +159,19 @@ public class RulingCleaningService {
private Rulings extractVerticalAndHorizontalRulingLines(List<Ruling> rulings) { private Rulings extractVerticalAndHorizontalRulingLines(List<Ruling> rulings) {
List<Ruling> vrs = new ArrayList<>(); List<Ruling> vrs = new ArrayList<>();
for (Ruling vr : rulings) { List<Ruling> hrs = new ArrayList<>();
if (vr.vertical()) { for (Ruling r : rulings) {
vrs.add(vr); if (r.vertical()) {
vrs.add(r);
}
if (r.horizontal()) {
hrs.add(r);
} }
} }
List<Ruling> hrs = new ArrayList<>();
for (Ruling hr : rulings) {
if (hr.horizontal()) {
hrs.add(hr);
}
}
return new Rulings(vrs, hrs); return new Rulings(vrs, hrs);
} }
private Ruling createRuling(float tableCellX0, float tableCellX1, float tableCellY0, float tableCellY1) {
float x0 = tableCellX0;
float x1 = tableCellX1;
float y0 = tableCellY0;
float y1 = tableCellY1;
if (x1 < x0) {
x0 = tableCellX1;
x1 = tableCellX0;
}
if (y1 < y0) {
y0 = tableCellY1;
y1 = tableCellY0;
}
return new Ruling(new Point2D.Float(x0, y0), new Point2D.Float(x1, y1));
}
private record Rulings(List<Ruling> verticalLines, List<Ruling> horizontalLines) { private record Rulings(List<Ruling> verticalLines, List<Ruling> horizontalLines) {
} }