Why OCR digitizes text from images in Computer Vision - Model Pipeline Impact

Model Pipeline - Why OCR digitizes text from images

OCR (Optical Character Recognition) changes pictures of text into real text that computers can read and use. This helps us search, edit, and store text from images easily.

Data Flow - 6 Stages

1Input Image

1 image (e.g., 600 x 400 pixels, grayscale)→Load image containing text→1 image (600 x 400 pixels, grayscale)

Photo of a printed page with letters and numbers

↓

2Preprocessing

1 image (600 x 400 pixels, grayscale)→Convert to grayscale, remove noise, adjust brightness→1 cleaned image (600 x 400 pixels, grayscale)

Clearer image with less background noise

↓

3Text Detection

1 cleaned image (600 x 400 pixels, grayscale)→Find areas likely containing text→Multiple text regions (e.g., 5 boxes)

Boxes around words or lines in the image

↓

4Character Segmentation

Text region images (varied sizes)→Split text regions into individual characters→Multiple character images (e.g., 50 characters)

Small images each containing one letter or number

↓

5Character Recognition

Character images (28 x 28 pixels each)→Use ML model to identify each character→Sequence of characters (e.g., 'HELLO123')

Predicted letters and numbers from images

↓

6Postprocessing

Sequence of characters→Correct errors, format text→Clean text string

'HELLO 123' as editable text

Training Trace - Epoch by Epoch

Loss
1.2 |****
1.0 |***
0.8 |**
0.6 |**
0.4 |*
0.2 |*
0.0 +----------------
      1 2 3 4 5 Epochs

Epoch	Loss ↓	Accuracy ↑	Observation
1	1.2	0.45	Model starts learning basic character shapes
2	0.8	0.65	Recognition accuracy improves as model learns
3	0.5	0.80	Model correctly identifies most characters
4	0.3	0.90	Loss decreases steadily, accuracy nears 90%
5	0.2	0.94	Model converges with high accuracy

Prediction Trace - 6 Layers

Layer 1: Input Image

Layer 2: Preprocessing

Layer 3: Text Detection

Layer 4: Character Segmentation

Layer 5: Character Recognition

Layer 6: Postprocessing

Model Quiz - 3 Questions

Why does OCR preprocess the image before detecting text?

A To remove noise and improve text clarity

B To add colors to the image

C To increase image size

D To convert text into numbers

Key Insight

OCR works by turning images into clear text through steps that clean the image, find text areas, split characters, and recognize them. Training improves the model's ability to read characters accurately, making text from images usable for computers.