Preliminary information


SmartApplets

SmartApplets are application related image processing libraries. Once loaded these can immediately be run. System adaptation will be effected by a graphical configuration. SmartApplets are individually adapted to compliant hardware and optimally use the potential of the according onboard vision processor.

Thematic focuses are combined in SmartApplets families, which are industry-, application- or technology-based oriented. SmartApplets are continuously enhanced. Hereby the number of families and the scope of each package will change.

SmartApplets provides a high quality pre-stage of Machine Vision applications. It reduces the load of host CPU by a complete outsourcing on the onboard FPGA. SmartApplets doesn’t cover single processing steps but a complete chain of concerted function blocks, from the acquisition over sensor corrections and image enhancements to special functions. Hereafter the images can the post-processed by embedding in own applications or transferring in 3rd party Machine Vision software.

Overview of SmartApplets family (state of the art, no responsibility is taken for the correctness of this information):

Smart Applets Familiy Binarization

Board
Img. Size
CamPort
Bitdepth
Sensor
LUT
Shading
Noise
Binarization
Morphology
Segm.
# DMA
Bandwidth
CamClk
mE4
Max.
.
.
Input
Output
Tap Corr.
8/12 bit
1D
Median
Thres.
Adapt.
Open/Close
Blob
.
Per cam
DMA/ch
Max.
CL/GE
[Px]
#
Type
[bit]
[bit]
.
.
mult+sub
3*3
.
Kernel
#
Kernel
Filter
#
[MP/s]
[MB/s]
[MHz]
Area Scan
Bin_BAG8
VD1
4k*4k
1
BASE
12
8
no
yes
no
yes
yes
32*32
2
5*5
n/a
2
170
250
85
Bin_BAG12
VD1
4k*4k
1
BASE
12
12
no
yes
no
no
yes
32*32
1
5*5
n/a
2
170
250
85
Bin_FAG8
VD1
4k*4k
1
FULL
8
8
no
yes
no
no
yes
32*32
n/a
2
500
250
85
Bin_BAG8
VD4
4k*4k
1
BASE
12
8
no
yes
no
yes
yes
64*64
2
7*7
n/a
4
170
250
85
Bin_BAG8_dual1
VD4
4k*4k
2
BASE
12
8
no
yes
no
no
yes
32*32
n/a
2
170
250
85
Bin_BAG8_dual2
VD4
4k*4k
2
BASE
12
8
no
yes
no
no
yes
no
2
7*7
n/a
2
170
250
85
Bin_BAG12
VD4
4k*4k
1
BASE
12
12
no
yes
no
yes
yes
64*64
2
7*7
n/a
2
170
250
85
Bin_FAG8
VD4
4k*4k
1
FULL
8
8
no
yes
no
yes
yes
64*64
1
3*3
n/a
2
850
750
85
Bin_GAG8
VQ4
4k*4k
1
GigE Vis.
12
8
no
yes
no
yes
yes
64*64
2
7*7
n/a
2
128
128
n/a
Bin_GAG12
VQ4
4k*4k
1
GigE Vis.
12
12
no
yes
no
yes
yes
64*64
2
7*7
n/a
2
128
128
n/a
Bin_GAG12_dual
VQ4
4k*4k
2
GigE Vis.
12
12
no
yes
no
yes
yes
32*32
1
7*7
n/a
2
128
128
n/a
Line Scan
Bin_BLG8
VD1
8k*65k
1
BASE
8
8
no
yes
no
no
yes
32*32
1
5*5
n/a
2
170
250
85
Bin_BLG12
VD1
4k*65k
1
BASE
12
12
no
yes
no
no
yes
32*32
1
5*5
n/a
2
170
250
85
Bin_FLG8
VD1
4k*65k
1
FULL
8
8
no
yes
no
no
yes
32*32
1
5*5
n/a
2
170
250
85
Bin_ BLG8
VD4
16k*65k
1
BASE
8
8
yes
yes
yes
no
yes
64*64
2
7*7
n/a
2
170
250
85
Bin_BLG12
VD4
16k*65k
1
BASE
12
12
yes
yes
yes
no
yes
32*32
1
7*7
n/a
2
170
250
85
Bin_MLG8
VD4
16k*65k
1
MEDIUM
8
8
yes
yes
*
no
yes
64*64
2
7*7
n/a
2
340
500
85
Bin_FLG8_2
VD4
16k*65k
1
FULL
8
8
yes
yes
*
no
yes
64*64
2
7*7
n/a
2
340
500
85
Bin_FLG8
VD4
16k*65k
1
FULL
8
8
yes
yes
*
no
yes
no
1
5*5
n/a
2
850
500
85
Bin_FLG8_3
VD4
8k*65k
1
FULL
8
8
yes
yes
no
no
yes
32*32
1
3*3
n/a
2
850
500
85

* Half shading resolution


Smart Applets Family Segmentation

Board
Img. Size
CamPort
Bitdepth
Sensor
LUT
Shading
Noise
Binarization
Morphology
Segm.
# DMA
Bandwidth
CamClk
mE4
Max.
.
.
Input
Output
Tap Corr.
8/12 bit
1D
Median
Thres.
Adapt.
Open/Close
Blob
.
Per cam
DMA/ch
Max.
CL/GE
[Px]
#
Type
[bit]
[bit]
.
.
mult+sub
3*3
.
Kernel
#
Kernel
Filter
#
[MP/s]
[MB/s]
[MHz]
Area Scan
Blob_BAG8
VD1
4k*4k
1
BASE
12
8
no
yes
no
yes
yes
no
2
7*7
Full
2
170
200
85
Blob_BAG12
VD1
4k*4k
1
BASE
12
12
no
yes
no
yes
yes
no
2
7*7
Full
2
170
200
85
Blob_FAG8
VD1
2k*4k
1
FULL
8
8
no
yes
no
no
yes
no
no
no
Full
2
850
200
85
Blob_BAG8_dual
VD4
4k*4k
2
BASE
12
8
no
yes
no
yes
yes
no
2
5*5
Full
4
170
250
85
Blob_BAG8
VD4
4k*4k
1
BASE
12
8
no
yes
no
yes
yes
32*32
2
7*7
Full
2
170
250
85
Blob_BAG12
VD4
4k*4k
1
BASE
12
12
no
yes
no
yes
yes
32*32
2
7*7
Full
2
170
250
85
Blob_FAG8
VD4
2k*4k
1
FULL
8
8
no
yes
no
yes
yes
32*32
1
3*3
Full
2
850
750
85
Blob_GAG8
VQ4
4k*4k
1
GigE Vis.
12
8
no
yes
no
yes
yes
32*32
2
7*7
Full
2
128
128
n/a
Blob_GAG12
VQ4
4k*4k
1
GigE Vis.
12
12
no
yes
no
yes
yes
32*32
2
7*7
Full
2
128
128
n/a
Blob_GAG8_quad
VQ4
4k*4k
4
GigE Vis.
8
8
no
yes
no
no
yes
no
no
no
Light
4
125
125
n/a
Blob_GAG8_dual
VQ4
4k*4k
2
GigE Vis.
8
8
no
yes
no
yes
yes
no
2
7*7
Full
2
125
125
n/a
Blob_GAG12_dual
VQ4
4k*4k
2
GigE Vis.
12
12
no
yes
no
yes
yes
no
2
7*7
Full
2
125
125
n/a
Line Scan
Blob_BLG8
VD1
16k*65k
1
BASE
8
8
yes
yes
yes
no
yes
no
1
5*5
Full
2
170
200
85
Blob_BLG12
VD1
16k*32k
1
BASE
12
12
yes
yes
yes
no
yes
no
no
no
Full
2
170
200
85
Blob_MLG8
VD1
16k*65k
1
MEDIUM
8
8
yes
yes
yes
no
yes
no
1
7*7
Full
2
250
200
62,5
Blob_MLG12
VD1
16k*65k
1
MEDIUM
12
12
yes
yes
yes
no
yes
no
no
no
Full
2
250
200
62,5
Blob_ BLG8_dual
VD4
16k*65k
2
BASE
8
8
yes
yes
yes
no
yes
no
no
no
Full
2
170
250
85
Blob_BLG8
VD4
16k*65k
1
BASE
8
8
yes
yes
yes
no
yes
32*32
2
7*7
Full
2
170
250
85
Blob_BLG12
VD4
16k*65k
1
BASE
12
12
yes
yes
yes
no
yes
32*32
1
3*3
Full
2
170
250
85
Blob_MLG8
VD4
16k*65k
1
MEDIUM
8
8
yes
yes
*
no
yes
32*32
2
5*5
Full
2
340
500
85
Blob_MLG12
VD4
16k*65k
1
MEDIUM
12
12
yes
yes
*
no
yes
no
2
7*7
Full
2
340
500
85
Blob_FLG8
VD4
16k*65k
1
FULL
8
8
yes
yes
*
no
yes
no
1
7*7
Full
2
640
750
85

* Half shading resolution

Abbreviations:

VD1 = microEnable IV VD1-CL
VD4 = microEnable IV VD4-CL
VQ4 = microEnable IV VQ4-GE

Example:

The example “Blob_BAG8” for microEnable IV VD4-CL demonstrates the workflow within a SmartApplet by block diagrams. The image acquisition (green) functions support grayscale BASE configuration cameras with a bit depth up to 12bit. A subsequent Lookup Table function (yellow) cares of image enhancements by gamma diagrams, contrasts, brightness etc. A median filter reduces an image noise. A binarization (orange) is processed with a configurable global threshold and an adaptive method with local threshold in a 32*32 pixel surrounding. After selection of the binarization method (SW), two-level morphology (yellow) with serial open-close function with 7*7 pixel matrix size, eliminates small objects. The Blob function (light blue) segments the image in binary objects. The object filter (blue) allows the passing of objects with certain properties by configuration of the feature extraction. By two DMA transfers, the SmartApplets outputs an object list with coordinates of bounding boxes, surface area, contour length and other object properties, and a grayscale image. The source of the image, before binarization or after the morholohic processing, can be selected by setting the according switch (SW).

Available models:

microEnable IV VD1-CL
microEnable IV VD4-CL
microEnable IV VQ4-GE

Above listed models are pre-licensed for “SmartApplets Base” at delivery and can be immediately loaded. “SmartApplets Base” covers applets from the Base function libraries. The usage of the SmartApplets from Extended function libraries (“SmartApplets Extended”) requires an additional license code.

language change language change top of page top of page
Suche