Date post: | 17-Jan-2016 |
Category: |
Documents |
Upload: | augustus-beasley |
View: | 241 times |
Download: | 1 times |
Machine Vision
Introduction to Using Cognex DVT Intellect
Intellect
Use 515 cameras which have CMOS greyscale sensors 640 x 480 pixels.
CCD are more precise and has readers and flaw detection but costs more!
Download from: • http://www.sigavision.com/siga_downloads.html
Intellect Menus
current camera
camera IP address
current working
pane
current software version
image ID display
Intellect shares many menus commonly found in other Windows applications like “File”, “Edit”, “View”, “Window”, and “Help”
Some menu items are unique to Intellect like “Image”, “System”, “Product”, and “System”
Tool Bars
Image Toolbarshow
previous image
(emulation)
show next image
(emulation)
start image cycle
show last inspected
image
stop image cycle
actual size
zoom to fit
zoom in
focus assist
record inspected
images
zoom out
image assist
method selection
enable image assist
set image sequence
Focus assistFocus assist – returns a score (sharpness, intensity and contrast), the higher the better
Image assistImage assist – automatically adjust gain and exposure time
System Toolbar
open Network Explorer
display all camera
connections
open system, product, or image
disconnect from
camera
save system to camera’s
flash
apply white
balance
undo toggle trigger method
(internal/external)
redo run/stop inspection
Product Toolbar
cut product
copy product paste
product
allows product
selection
delete product
toggle antiblooming
activate external LED ring lights
save to camera’s
flash
set exposure
(ms)
create new product
Product Management
Used in conjunction with the properties window to:- Rename Product- Put in a description- Assign Coordinate Transform or Global Scaling- Change the Digital ID of a Product (for digital input selection of product to use)
Tool Toolbar
paste tool
delete tool show tool
properties
allows tool selection
reposition tools
copy tool
cut tool
Toolbox Pane
Camera limitations
510 & 515 do not do Readers or Flexible Flaw detection.
Manual also states 510 does not do positioning or counting but this is probably incorrect!
Pre-processing (Filtering)
Enhance Image Features• Edges
• Defects
Reduce Noise Increase Contrast (image domain) Subtract Background Feature Adds processing time!
Types of Filter Gradient Low pass High pass Sharpen Median Kirsh Convolution
Erode Dilate Open Close Fill Holes FOV Balance Image Domain
Subtract Dark Image Color Inverse Log Edge Aggregate
Convolution
Example from ES372:Spatial-domain: Convolution Mask
Best shown by example:
Convolution mask (3x3)
Image
New value is calculated as:
29111111111
)1201201201201100120120120120(
Sum of matrixSum of product of
both matrices
Positioning Tools
You can find positioning tools by clicking the Positioning header within the Toolbox
There are four positioning tools from which to choose
searches along a line for an edge, feature, or pattern
searches an area for the location of trained feature(s)
searches along an area for a specific edge
finds the center position and the radius of a circle
Counting Tools
CountAlong Line
EdgeCount
Count inArea
PixelCounting
ReplicatorTool
PatternCount
BlobPatternMatch
ObjectLocate
Results used for a Source of Positions to
place another tool, multiple times
FeatureCount
Learn Pixel Intensities
Object locate will probably give the best results
Identification tool
Identifies the part by using multiple leaned models by either Blob properties or Colour properties
Scripting
Like JavaScript Always make comments about what you
are trying to do Reserved words:
• They have special meaning
• Ex. if and PASS End almost all lines with ;
Coordinate Systems
Allows for transforms to be set up for Coordinate Transformations, lens distortion and Tool scaling.
Under systems menu.Great for traditional
robot/camera/world calibration
Camera
(grid) Frame
Robot Frame
+X
+Y
(0,0)
+X
+Y
(0,0)
Coordinate Transformation
Coordinate Systems
Place DVT Target under camera Click “Calibrate” to create a Transform Choose this Transform for your Product