CDIA Image Processing 5 — Questions and Answers
Question 1: In document image processing, what is 'zone OCR' used for?
- Applying OCR only to predefined regions of a document rather than the full page (Correct answer)
- Scanning zones of a physical document sequentially
- Compressing specific areas of an image more aggressively
- Identifying security zones in classified documents
Correct answer: Applying OCR only to predefined regions of a document rather than the full page
Zone OCR restricts text recognition to predefined regions (zones) on a page, improving speed and accuracy for structured forms and templates.
Question 2: Which image artifact is introduced when a scanner's CCD sensor has uneven sensitivity across its width?
- Skew
- Vertical streaking or banding (Correct answer)
- Moiré patterns
- Bleed-through
Correct answer: Vertical streaking or banding
Uneven CCD sensitivity causes vertical streaks or banding along the scan direction, appearing as light or dark vertical lines in the scanned image.
Question 3: What is the CDIA-recommended minimum resolution for scanning documents intended for long-term archival?
- 72 DPI
- 150 DPI
- 300 DPI (Correct answer)
- 600 DPI
Correct answer: 300 DPI
300 DPI is the widely accepted minimum for archival-quality document scanning, balancing image fidelity with manageable file sizes.
Question 4: What is the function of an image 'normalization' step in a document processing pipeline?
- Converting all documents to the same file format
- Standardizing image properties such as resolution, bit depth, and orientation for consistent downstream processing (Correct answer)
- Removing all metadata from scanned images
- Applying the same compression to every document
Correct answer: Standardizing image properties such as resolution, bit depth, and orientation for consistent downstream processing
Normalization standardizes key image properties across a batch so that subsequent processing steps (OCR, indexing, storage) receive consistent input.
Question 5: In the context of document imaging, what is a 'confidence score' in OCR output?
- A measure of scanner hardware reliability
- A numeric value indicating how certain the OCR engine is about each recognized character or word (Correct answer)
- The compression quality level applied to the image
- The percentage of the document successfully scanned
Correct answer: A numeric value indicating how certain the OCR engine is about each recognized character or word
OCR confidence scores reflect the engine's certainty for each recognized character or word, allowing low-confidence results to be flagged for human review.
Question 6: Which approach is used to handle mixed-orientation pages (portrait and landscape) in a batch scanning workflow?
- Manual operator rotation of each page before scanning
- Automatic orientation detection and rotation using text-direction analysis (Correct answer)
- Scanning all pages at 600 DPI to compensate
- Converting all pages to a single standard orientation before OCR
Correct answer: Automatic orientation detection and rotation using text-direction analysis
Automatic orientation detection analyzes text line direction and character patterns to determine and correct page orientation without manual intervention.
Question 7: What is the purpose of 'image segmentation' in document analysis?
- Dividing a document page into meaningful regions such as text blocks, images, and tables (Correct answer)
- Splitting a large image file into smaller chunks for faster transfer
- Separating the foreground from background color layers
- Identifying individual pages in a multi-page document
Correct answer: Dividing a document page into meaningful regions such as text blocks, images, and tables
Image segmentation partitions a page into logical regions (text, graphics, tables, headers) to enable targeted processing and accurate layout analysis.
CDIA Image Processing 5
In document image processing, what is 'zone OCR' used for?