Table Instance - Read Empty Cells

Capture the contents of empty cells within a Table Instance.

Remarks

The Read Empty Cells command is designed to automatically populate table cells that remain empty after initial extraction or manual adjustment of table headers and rows. This command is especially useful in scenarios where cell boundaries are well-defined by the table structure, but some cells lack extracted values due to OCR limitations, ambiguous content, or incomplete extraction logic.

How It Works

• The command iterates through each row and cell in the target Table Instance.
• For every cell that is blank, it determines the geometric region associated with the cell using the corresponding column header and row boundaries.
• It then extracts the text content from this region using OCR or other extraction methods.
• If any content is found, the cell is populated and validated automatically.

Usage Guidance

• Use this command after configuring or adjusting table headers and rows, especially when some cells are left empty due to extraction gaps.
• It is ideal for post-processing tables where cell regions are known, but not all data was captured during initial extraction.
• The command helps ensure data completeness and reduces manual data entry by leveraging geometric cell boundaries.

Extraction Workflow

1. The command checks each row and cell for blank values.
2. For each empty cell, it locates the corresponding header and computes the cell's geometric region.
3. Text is extracted from the region and, if present, assigned to the cell.
4. The cell is validated to ensure the extracted value meets configuration requirements.

Best Practices

• Run this command after header and row adjustments to maximize data capture from well-defined table regions.
• Use in combination with other extraction commands to fill in missing data and improve overall table accuracy.
• Review extracted values for quality, especially in documents with poor image quality or complex layouts.

For more information, see documentation for Table Instance, Table Row Instance, and Table Cell Instance.

See Also

Table Instance