Smart Gui - Algorithm Training Material_rev1.pdf - 第110页

Copyright © ViT rox All Rights Reserved . S-type Algorithm - Body Mode 1 10 Presence Mode Description Classic This is based on examining the grey level of a defined area of the component. Colour MVC Its reliability is ba…

109

Component Body Inspection

S-type Algorithm - Body Mode

110

Presence Mode Description

Classic This is based on examining the grey level of a defined area of the component.

Colour MVC Its reliability is based upon board color and contrast to component color. The

classifier is based on inspecting 3 populated boards, and 3 bare boards. This

method is not ideal for small dark components, with low contrast to the board

color, like diodes. This cover in SLA training module.

PRMVC This is based on 3 populated boards, and 3 bare reflowed boards. This

requires extra CAD generation to create pads information. This cover in SLA

training module.

Voting This takes all of the classifiers into consideration. The algorithm then uses a

weighted score from each classifier and votes for presence/absence. This is a

classifier that is based on the shape profile of the component. This cover in

SLA training module.

In most cases, Classic is sufficient for

determining presence on s-types. If you

have a bare reflowed board available, you

can use the Voting classifier and train the

SLAs.

S-type Algorithm - Body Mode (Classic)

1. Body Presence consists of 2D detection & 3D Detection.

2. 2D Classic Presence/Absence takes the average grey scale for the red lead boxes and subtracts the average grey scale for the green

boxes to give a presence score.

3. 3D Classic takes the inspected average lead height in the red box and compares it to the presence threshold pre set.

4. This is the recommended solution for detecting presence/absence for s-type, components.

111