Transcript Practice

Practice
Introduction of DT Vision
• Open image
• Image type
• Status Bar contains: Image file name,
Image type, coordinate, pixel value
• ROI type
– Draw Active (RED color) or inactive ROI
– Delete ROI
Continue
• Using the arithmetic tool
–
–
–
–
Add
Subtract
Multiply
Divide
•
• Blob analysis:
– What is a blob? : Number of connected pixels
within a specified threshold range
– Find, count, measure all blobs within the ROI
Continue
– Test different blob setting (Check the
results caused by different setting)
– save blob setting for later processing
– DDE : transfer statistical processing
result to Excel file
• Calibration—
– Why we need calibration ?
• Color plane
– Add Histogram based on RGB
• Contour classifier
Continue
•
Contrast enhancement
–
–
•
Stretching LUT
Histogram equalization
Edge finder tool
–
1.
2.
3.
Stages
Active threshold
Auto threshold
Binary
Preprocessing
• Image Acquisition
– Image format
– Transform for different format
• RGB to HIS or HUV
– Pixel type
• Histogram analysis of the fetched
image
– Identify uniform or non-uniform graylevel distribution
Preprocessing (Continue)
• Contrast enhancement
– Noise removal
• Median filter
• Max filter
• Min filter
– Sharpening for blurring edge
• Laplacian
• Sobel operator
– Smoothing for sharp edge
– Color enhancement
• Histogram euqalization
Segmentation
• Thresholding based on histogram: bi-level,
multi-level
• Region based : based on gray level
similarity
– Region growing
– Split and merge
• Edge based
– Point detection
– Line detection
– Edge detection
Feature extraction
• Feature representation
– Invariant to translation, rotation, and
scaling
– Texture
– Geometry
Classification
• Artificial Intelligence
–
–
–
–
Neural Network
Fuzzy C- Means
Genetic algorithm
Tabu Search