Eyesweb Compendium

  • October 2019
  • PDF

This document was uploaded by user and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this DMCA report form. Report DMCA


Overview

Download & View Eyesweb Compendium as PDF for free.

More details

  • Words: 5,314
  • Pages: 56
EyesWeb Compendium 2004

1

EyesWeb Compendium 2004

Index 1

Introduction........................................................................................................... 3 1.1 What is Eyesweb? ............................................................................................. 3 1.2 Getting Started with EyesWeb............................................................................. 4 1.3 Intel OpenCV .................................................................................................... 5 1.4 Workspace presentation ..................................................................................... 6 1.5 What is a Patch? ............................................................................................... 8 1.6 What is a Block? ............................................................................................... 9 1.7 Tutorials ........................................................................................................ 12 1.8 Asking questions ............................................................................................. 13 1.9 Newsgroups ................................................................................................... 14 2 Block Documentation ............................................................................................ 15 2.1 Generic.InputSwitch ........................................................................................ 15 2.2 Generic.PeriodBang ......................................................................................... 17 2.3 Generic.SnapShot ........................................................................................... 18 2.4 Generic.Switch................................................................................................ 19 2.5 Generic.Trigger ............................................................................................... 20 2.6 Generic.Queue................................................................................................ 21 2.7 Generic.UnaryOp............................................................................................. 22 2.8 Imaging.Blobanalysis.ExtractRegions ................................................................. 23 2.9 Imaging.Blobanalysis.TrackRegions ................................................................... 25 2.10 Imaging.Conversion.ColorToGray.................................................................... 26 2.11 Imaging.Blobanalysis.GrayToColor .................................................................. 27 2.12 Imaging.Draw.2PointElement ......................................................................... 28 2.13 Imaging.Filters.NonLinearFilter ....................................................................... 29 2.14 Imaging.Input.FilledFrame ............................................................................. 30 2.15 Imaging.Input.MultimediaFileRead .................................................................. 32 2.16 Imaging.Operations.DyadicArithmeticOp .......................................................... 34 2.17 Imaging.Operations.ExcludeAreas ................................................................... 35 2.18 Imaging.Operations.MonadicLogicalOp............................................................. 36 2.19 Imaging.Operations. MorphologicalOp ............................................................. 37 2.20 Imaging.Operations.Threshold........................................................................ 38 2.21 Imaging.Output.Display................................................................................. 39 2.22 Math.Matrix.BinaryOp ................................................................................... 40 2.23 Math.Matrix.Concatenate ............................................................................... 41 2.24 Math.Matrix.DomianConv............................................................................... 42 2.25 Math.Matrix.Extract ...................................................................................... 43 2.26 Math.Matrix.GetEntr...................................................................................... 44 2.27 Math.Matrix.Input.Generator .......................................................................... 45 2.28 Math.Scalar.BinaryOp.................................................................................... 46 2.29 Math.Scalar.BinaryOpLogical .......................................................................... 47 2.30 Math.Scalar.RunningOp ................................................................................. 49 2.31 Math.Scalar.UnaryOpLogical........................................................................... 50 2.32 Math.Scalar.ConstOp .................................................................................... 51 2.33 Math.Scalar.ThresCrossing............................................................................. 53 2.34 MoCap.Item.Extract ...................................................................................... 54 2.35 MotionAnalysis.Cues.SMI ............................................................................... 56

2

EyesWeb Compendium 2004

1 Introduction 1.1

What is Eyesweb?

In short Eyesweb is a Microsoft Windows based tool for creating interactive digital multimedia applications. The tool enables the possibility to form digital sound and images real-time, through use of various Human-Computer Interactions (HCI). These interactions include but are not limited to: Object Identification, Segmentation and Recognition, Face Recognition, Gesture Recognition and Motion Tracking. The Eyesweb tool is used by artists all over the world for exploring the universe of digital music and visuals.

Fig.1. Pics of the Eyesweb Platform Why is Eyesweb of relevance to you? Well have you ever wondered how the professionals make and use the Bluescreen technique as seen in e.g. Star Wars and Lord of the Rings? Ever wondered how a computer can recognise a fingerprint at the police station? Ever wondered how some surveillance systems are able to track moving objects automatically even at night? Well all these things are done through digital image processing and can be achieved in Eyesweb as well. Therefore exploring Eyesweb will introduce you to the theory that lies beneath a lot of processes that some today take for granted. Through your work with Eyesweb you will gain a better understanding of basic digital image and sound processing theory. The knowledge obtained in this course will then open ones eyes in regards to many other programs and applications and you will see these in totally new ways.

3

EyesWeb Compendium 2004 Eyesweb is a non-profit open software platform developed by Lab. InfoMus - DIST University of Genoa, Italy. For more information from the creators themselves see: http://www.infomus.dist.unige.it/eywmain.html .

Eyesweb is based on the Intel OpenCV (Open Source Computer Vision Libaray). The Intel OpenCV is a collection of program code (C functions, few C++ classes and popular algorithms) obtained and contributed by researchers all over the world. The Intel OpenCV may be a bit hard to understand for non-programmers but often contributes to “the understanding

of

things”

through

its

documentation.

For

more

info

see:

http://www.intel.com/research/mrl/research/opencv/

1.2

Getting Started with EyesWeb

To get properly started with the Eyesweb tool, the first thing to do is of course to install the program itself. Since there is a lot of different versions and plug-ins out there, it is crucial that you install the program in the following manner:

1.

Download and install the Eyesweb v.3.3.0 Platform from: ftp://infomus.dist.unige.it/Evaluate/Eyesweb/Version_3.3.0/EyesWeb_full_3.3.0.exe

2.

Download and install the Eyesweb v.3.3.0 Upgrade sp1 from: ftp://infomus.dist.unige.it/Evaluate/Eyesweb/Version_3.3.0/EyesWeb_upgrade_3.3.0sp1.e xe

3.

Download and install the Eyesweb v.3.0.2 Motion Analysis Add-on from: ftp://infomus.dist.unige.it/Evaluate/Eyesweb/Version_3.1.0/EyesWeb_addons_3.0.2.exe

4.

And finally download and ALWAYS USE the Eyesweb 3 documentation from: ftp://infomus.dist.unige.it/Evaluate/Eyesweb/version_3.0.0/EywDocs.zip When having completed these steps Eyesweb are ready for use.

4

EyesWeb Compendium 2004 1.3

Intel OpenCV

If you are keen on some more “hard-core” documentation, then the Intel OpenCV is a open possibility. Note: The Intel OpenCV is NOT mandatory for the AP course (but it is advised to try it out). To install Intel OpenCV, take the following steps:

1.

Goto: http://sourceforge.net/projects/opencvlibrary/

2.

Under latest released files find opencv-win and press ‘Download’

3.

Here find the ‘OpenCV_b3.1.exe’. Download and install this file. It’s 17,5mb so have patience.

4.

After installing just open the documentation of the OpenCV in the start menu. Here you’ll find a vast amount of useful stuff.

5

EyesWeb Compendium 2004 1.4

Workspace presentation

A first time user of the Eyesweb tool would be wise to segment the interface into 4 major areas as shown in Fig.1 below.

Fig.2 The Eyesweb Workspace

Main Workarea This is the area where a developer creates his Eyesweb Patches. Objects are moved around here with the drag n’ drop method as we know it from windows. The developer navigates around this area with the standard horizontal and vertical scrollbars or with the Patch Navigator window.

6

EyesWeb Compendium 2004 Tools The Toolbars of Eyesweb are, as in any other program, shortcuts to the different operations of the program. It is advised to activate ALL toolbars under the ‘View’ menu if allowed by ones screen-size/resolution. Playing around with the different toolbars is an efficient way to learn their functionality and should be done by a first time user.

Libraries The libraries of Eyesweb are a structured index of the different blocks that the program offers. It is advised to study the 15 different root folders to learn what lies beneath their titles. In this way a new user will get an initial overview and hence a better understanding of the programs structure and limits before commencing the actual development.

Commandline As in most other programs the commandline is a text based feedback to the user. The commandline is most useful to identify blocks, to debug patches through is error reporting and much more. It would be a good idea to implement usage of the commandline in ones workflow at an early point, if one is new to Eyesweb.

7

EyesWeb Compendium 2004 1.5

What is a Patch?

A Patch is a term used to define an Eyesweb file. In terms of Eyesweb a patch is a structured network of blocks that channels and manipulates a digital input dataflow resulting in a desired output dataflow (see Fig.2 below). This manipulation can be done automatically or through real-time interaction from a user (e.g. by motion tracking). The build-up of a patch in many ways resembles that of an object oriented program (a network of clusters, classes and objects). A patch could therefore be interpreted as ‘a small program’.

Fig.3 Patch I/O Illustration Structuring a patch is often an almost undoable job since patches easily grow to enormous proportions. It is therefore strongly advised that you build-up your patch in small clusters (gatherings of blocks that have relevance to each other) and that you skilfully administrate the connections between blocks. In this way documenting your work for others to read is also easier.

Fig.4 Other outputs generated from the input of Fig.2

8

EyesWeb Compendium 2004 1.6

What is a Block?

If a patch is a structured network of blocks, then what is a block? As mentioned in the last section, a patch can be perceived as an object oriented program. If a patch is a program then the blocks should be considered as the individual classes and functions of the program. Each block is then an instance of a class and has a specific function that contributes to the final output of the patch. So blocks are simply named ‘blocks’ because you use them to build something of a bigger whole.

Remember that you can always identify a block by hovering the cursor over it and reading the commandline OR by right clicking it and then choosing ‘Locate’.

Fig.5 Block I/O illustration

Above (Fig.4) you see an illustration of a typical block. As the patch a block of course has an input and an output. Each block also has a range of attributes that we’ll now quickly go through.

Params: The ‘params’ are the main parameters of a block that the user can set. These ‘params’ defines or refines the function of the specific block. If in doubt about the function of a block, remember that you can always find the function of a specific block in the Eyesweb

9

EyesWeb Compendium 2004 documentation! You can see the ‘params’ of a block by double clicking it or by right clicking it and then choosing ‘set params’.

Extended Params: Though they are seldom used the ‘extended params’ of a block holds one useful function and that is the deactivation-button. If ever wondering what a specific block does in a patch, it is always a good idea to try to deactivate it. In this way you will simply see through the output what the block does. You can see the ‘extended params’ of a block by double clicking it or by right clicking it and then choosing ‘set params’.

Exported Params: The ‘exported params’ node are ‘params’ that you can set while a patch is running. In other words these are ‘params’ that can be manipulated real-time. This can be done manually by the user or through the output of another block. This function has many advantages when making HIC applications in Eyesweb. You can see the ‘exported params’ of a block by double clicking it or by right clicking it and then choosing ‘set params’.

Input: The input node of a block is of course where the block gets its input dataflow. As in programming a block MUST get a specific input to give a specific output. Giving a block that expects a matrix input an int instead will e.g. result in error. If getting such errors ALWAYS check that you are giving a block the correct type of input. You can see what type of input a block expects by hovering the cursor over the blocks input node and then reading the commandline.

Output: As with the input node, the output node is of a certain type. You can learn this type in the same manner as with the input. To understand the output of a block it is a good idea to ‘read it’ by using the output block of the given library. If you e.g. want to read an output of type matrix, you’d have to find a math.matrix.output.display block under the ‘math’ library.

10

EyesWeb Compendium 2004 Note: It is a VERY good idea to try to read the output of a block if you are either uncertain of what it does OR are trying to identify an error.

Custom (only some blocks): Some blocks have some custom ‘params’ that can be set. These special ‘params’ are very block specific and are usually rather easy to figure out. If in doubt; play around with them, refer to the documentation or ask a teacher (in that order).

Block Info: If ever in doubt of what a specific block does, simply right-click the block and choose ‘block info’. You will then get a short description of the blocks function. If this does not help you understand the block, please refer to the block documentation in the ‘documentation’ dir under the Eyesweb folder in the ‘Start’ menu.

11

EyesWeb Compendium 2004 1.7

Tutorials

Besides the exercises presented in the Automatic Perception course, tutorials and examples for Eyesweb can be found else where. Under your Eyesweb installation folder you’ll find a directory named /Patches. This folder contains about 100 patches and test-movies free to use. So if you a keen on learning more about Eyesweb fast, try to investigate these example patches. You’ll find ,no doubt, that these patches are often a great reference if you need inspiration for a specific problem or project. So if you are determined on doing things yourself, this is a good place to start.

Further tutorials can also be found on the Eyesweb website under http://www.infomus.dist.unige.it/eywtutorials.html.

12

EyesWeb Compendium 2004 1.8

Asking questions

When asking questions, in the time to come, remember ALWAYS use the Medialogy Forum (http://medbb.cs.aue.auc.dk/). AP related questions send directly by mail to either teacher or teaching assistant simply won’t be answered. Why is this so important? Well if everybody uses the forum for questions as well as answers, the forum will automatically turn into a FAQ for Automatics Perception. Hence future generations of Medialogy will be able to use this FAQ and save their own as well as the teachers’ time. All in all a better solution for everybody.

Furthermore, to give answers as quickly and efficient as possible, we teachers urge you students to include the following information in your questions:

Question: What is the problem and what are you trying to achieve? Own Effort: What has already been researched and where are you now? Ideas: What do you THINK could help solve the problem? Work: Do you have an Eyesweb patch that shows your current state and progress?

Example: “Hey Teacher We, the people of group x, are trying to do some background subtraction on an input movie stream. We are currently stuck though since our output looks totally wrong. We’ve been playing around with the Boolean operators, have actually tried all of ‘em, but still aren’t getting anywhere. We have an Idea that we might be using the wrong block type. Does this hold any truth? We have attached our current patch version. Please have a look at it give some feedback. Best Group x”

13

EyesWeb Compendium 2004 1.9

Newsgroups

The creators of Eyesweb has also made newsgroups available for questions. To activate these simply find the ‘support’ dir under the Eyesweb folder in the ‘Start’ menu and click the newsgroup of interest.

The available groups are: •

eyesweb.announce General announcements (new releases, fixes, etc.) or availability of tools and utilities.



eyesweb.bugs Bug reports, workarounds.



eyesweb.discussion Exchange of ideas, proposals, uses of the system, etc.



eyesweb.install Discussion of issues related to EyesWeb install and setup procedures on different systems.



eyesweb.libraries Announcements and discussion of current libraries, new blocks and patches from the user community, etc.



eyesweb.patches Exchange of patches and related material.

14

EyesWeb Compendium 2004

2 Block Documentation 2.1 Generic.InputSwitch Input/Output Datatype:

Description: This block functions as a ‘redirector’ of data-stream. Hence this block is often used to control the dataflow of a patch. The InputSwitch block receives 2 inputs as default, but can receive any number of inputs. The block switches between inputs based on the int value of the external param. Example: In its default state the block will let input 0 pass until another int input is received through the external param.

At this point then block will then ‘Switch’ and let

input 1 pass if the external param input received is 1.

Parameters: Name

Type

Description

Working mode

Combo

Selects which of the inputs to send to the output.

Num inputs

Integer

The number of inputs. Selects the active input that can be used as sync.

Active input

Integer Min. value = 0

15

EyesWeb Compendium 2004 Can lose data

Boolean

16

EyesWeb Compendium 2004

2.2 Generic.PeriodBang

Input/Output Datatype:

Description: This block generates a periodic Bang output (bang signals with given time delays). TIP: Bang signals can be used to activate a command button.

Parameters: Name

Type

Period

Unit(ms)

Description The time interval in ms of the generated bangs.

17

EyesWeb Compendium 2004

2.3 Generic.SnapShot

Input/Output Datatype:

Description: This block is used to grab a single frame of a video stream when activated through its external param. Hence the name ‘Snapshot’. It provides as output the value stored in an internal buffer. TIP: Use the "Load" command to load the internal buffer with a new value (take a new snapshot). Activate the "Load" command with a button or Bang signal.

Parameters: Name

Type

Load

Command

Description Loads one image on command.

18

EyesWeb Compendium 2004

2.4 Generic.Switch Input/Output Datatype:

Description: Enables or disables the flow of data through a channel. When working in 'Switch' mode the block keeps the current status (enabled or disabled channel) until a new 'Switch' command is received through the external param. When working in 'Trigger' mode the channel is automatically disabled once data has passed through. TIP: Use the 'Trigger' command to enable the block again when in ‘Trigger’ mode for another data instance.

Parameters: Name

Type

Switch/Trigger

Command

Mode

Combo

On/off

Boolean

Description Activation command parameter. The current mode of the block. Can be either ‘Switch’ or ‘Trigger’. States the current state of the block

19

EyesWeb Compendium 2004

2.5 Generic.Trigger Input/Output Datatype:

Description: Propagates the input to an output, whenever a ‘Pulse’ command is received. The block resets itself after each trigger instance.

Parameters: Name

Type

Pulse

Command

Description Activation parameter

20

EyesWeb Compendium 2004

2.6 Generic.Queue Input/Output Datatype:

Description: Provides a queue of arbitrary length to delay the dataflow. Example: This block will grab and store the 4 frames back in time if Length is set to 4. TIP: Use this block if you want to ‘store’ previous frame data for further processing.

Parameters: Name

Type

Length

HasMin(0)

Description The length of the queue

21

EyesWeb Compendium 2004

2.7 Generic.UnaryOp Input/Output Datatype:

Description: This block does unary operations on images, sound buffers, scalars, and matrices. In others words ‘various operations on a single input’.

Parameters: Name

Type

Description The different operations that can be used in this block is: No operation Sum Subtraction

Operation

Combo

Inverse subtraction Mean Mult Division Abs Square

22

EyesWeb Compendium 2004

2.8 Imaging.Blobanalysis.ExtractRegions Input/Output Datatype:

Description: This block identifies multiple blobs of white pixels from a binary image. The regions are calculated in a top left first order. TIP: This Block is used in blob-tracking. To track and identify regions use the TrackRegions block linked to the output of this block.

Parameters: Name

Type

NumRegions

Interger

ScaleFactor

Interger

Thresh

Interger

CloseIterations

Interger

Description The number of maximum regions to extract. Number of times to reduce the size of the image for quantization. Threshold value applied to the reduced quantized image Number of times the close algorithm is applied to the thresholded image

23

EyesWeb Compendium 2004 Output

Boolean

Toggles the quantized image output

24

EyesWeb Compendium 2004

2.9 Imaging.Blobanalysis.TrackRegions Input/Output Datatype:

Description: Tracks and identifies multiple blobs, extracted by the ExtractRegions block. TIP: This Block is used in blobtracking.

Parameters: Name

Type

Reset

Command

Description ExportedByDefault

25

EyesWeb Compendium 2004

2.10 Imaging.Conversion.ColorToGray Input/Output Datatype:

Description: Converts a 24-bit color image to a 8-bit grey scale image. This results in a 1 channel image.

Parameters: Name

Type

Description Depth can be set to: The same as input or

Depth

Combo

1-bit 8-bit 16-bit

26

EyesWeb Compendium 2004

2.11 Imaging.Blobanalysis.GrayToColor Input/Output Datatype:

Description: Converts a 1 channel 8-bit grayscale image into a 3 channel 24-bit grey scale image.

Parameters: Name

Type

FractR

Double

Description Min. 0, if the red color should not be converted. Max. 1, if the red color should be converted. Min. 0, if the green color should not be

FractG

Double

converted. Max. 1, if the green color should be converted.

FractB

Double

Min. 0, if the blue color should not be converted. Max. 1, if the blue color should be converted.

27

EyesWeb Compendium 2004

2.12 Imaging.Draw.2PointElement Input/Output Datatype:

Description: Draws a Line or a Rectangle in the Input Image using an image and a matrix as inputs. This is typically used to identify blobs in blobtracking. TIP: The image input could be the original dataflow and the matrix input a matrix holding [x, y, width, height] of a tracked blob. This could result in a bounding rect around a moving object in a movie.

Parameters: Name

Type

Description Type(0) Line

Type

Combo Type(1) Rect

Set Color

Command

Sets the drawing color of the line or rect.

28

EyesWeb Compendium 2004

2.13 Imaging.Filters.NonLinearFilter Input/Output Datatype:

Description: Performs a nonlinear rank filtering on the input image. TIP: This block is very useful when doing noise suppression.

Parameters: Name

Type

Description The different filter types that can be used in this block is:

Filter type

Combo

Median Min Max Color median

Number of columns Number of rows

Integer

Specifies how many columns the filter is made of.

Integer

Specifies how many rows the filter is made of.

29

EyesWeb Compendium 2004

2.14 Imaging.Input.FilledFrame Input/Output Datatype:

Description: Generate a frame with the given dimensions ('Width' and 'Height' parameters). If 'Mode' is set to 'Color', then the output image is a 3 channel 24-bit image, with each of the channels set to the values specified by the 'Red', 'Green', and ''Blue' parameters, respectively. If 'Mode' is set to 'B/W', then the output is a 8-bit grey scale.

Parameters: Name

Type

Width

Integer

Height

Integer

Description The width of the output frames in pixels. Min. size = 0, max. size = 2000. The height of the output frames in pixels. Min. size = 0, max. size = 2000. Two options, 'Color' or 'B/W' mode.

Mode

Combo

If 'Colour' is chosen the output is a three channel image containing red, green and blue, if 'B/W' is chosen the output is a greyscale image.

Red

Integer

The amount of red in the output image. Min. = 0, max. 255

30

EyesWeb Compendium 2004

Green

Integer

Blue

Integer

Fps

Double

Pick_color

Command

The amount of green in the output image. Min. = 0, max. 255 The amount of blue in the output image. Min. = 0, max. 255 The amount of output frames pr. second. Min. 0.001, max. 100. Allows color selection via the Windows Color Picker Common Dialog.

31

EyesWeb Compendium 2004

2.15 Imaging.Input.MultimediaFileRead Input/Output Datatype:

Description: This block reads a video file from a given location (e.g. harddrive) and gives this signal as an output. It is usually activated with a command button through its external param input. The params of this block include framerate, looping, start/end frame and more. TIP: If your system is stalling, try lowering the framerate to free memory. To make a relative link to a movie clip remove path info from filename. Press the ‘SetOrigParam’ button after defining a movie path. This ‘auto’ sets the params of the block.

Parameters: Name

Type

Description

AviFileName

FileName

The avi-file used. The number of frames per second the movie

FrameRate

Double

plays. Min. fps. 0.001. Max. fps. 100.

StartFrame

Integer

The frame from where the movie should begin.

StopFrame

Integer

The frame where the movie should stop.

32

EyesWeb Compendium 2004 Reads the start frame, stop frame and frame rate SetOrigParams

Command

of the AVI file and changes the parameters of this block accordingly.

Continuos Output

Loop

Play, Pause, Stop

Realtime

When true, the block should continue to send Boolean

frames to the output when the block is paused or stopped.

Boolean

When false, the movie should rewind when the end of the file is reached.

Command

Boolean

When true, the block may skip frames if it´s overloadet to keep the framerate constant.

33

EyesWeb Compendium 2004

2.16 Imaging.Operations.DyadicArithmeticOp Input/Output Datatype:

Description: This Block performs a dyadic arithmetic operation on two input images. In others words ‘various operations on two inputs’.

Parameters: Name

Type

Description The different operations that can be used in this block is: Add

Operation type

Combo

Subtract Subtract inverse Subtract absolute Multiply Multiply scale

34

EyesWeb Compendium 2004

2.17 Imaging.Operations.ExcludeAreas Input/Output Datatype:

Description: This Block erases multiple rectangular areas with the specified background color. The areas are specified with a DisplayClick block or a dynamic matrix input. TIP: This block can be used to extract a ROI from an image stream. The matrix input defines which area to exclude/include.

Parameters: No declared parameters.

35

EyesWeb Compendium 2004

2.18 Imaging.Operations.MonadicLogicalOp Input/Output Datatype:

Description: Performs a monadic logical operation on a single input image. All operations are performed bitwise, i.e. between corresponding pixels. Tip: Using shifts is the same as dividing or multiplying by powers of 2.

Parameters: Name

Type

Description The different filters that can be used in this block is: Left shift

Operation type

Right shift

Combo

Not And Or Xor

Value

Integer

The number of places to shift by, or use boolean operator against.

36

EyesWeb Compendium 2004

2.19 Imaging.Operations. MorphologicalOp

Input/Output Datatype:

Description: Performs a morphological operation on the input image. Either erode, dilate, open or close. TIP: These blocks are very useful tools in noise suppression.

Parameters: Name

Type

Description The different operators that can be used is: Erode

Operation type

Combo

Dilate Open Close.

Num. Iterations

Integer

The number of iterations the morphological operator should be applied.

37

EyesWeb Compendium 2004

2.20 Imaging.Operations.Threshold

Input/Output Datatype:

Description: Thresholds the source image. Everything above the threshold is considered ‘foreground’ and everything under is considered ‘background’

Parameters: Name

Type

Description

Threshold

Integer

The value the threshold is set to.

38

EyesWeb Compendium 2004

2.21 Imaging.Output.Display

Input/Output Datatype:

Description: The Image.output display is used as an image output. The block takes the image from the image-library and displays it in a popup window.

Parameters: Name

Type

Interpolation

Combo

MinimizeWithG UI

Boolean

Description Indicates which Interpolation algorithm when the window and image size differ. Indicates whether the display window should be minimized with the Eyesweb application.

39

EyesWeb Compendium 2004 2.22 Math.Matrix.BinaryOp Input/Output Datatype:

Description: Performs the selected operation on two input matrices resulting in a new matrix.

Parameters: Name

Type

Description The different operations that can be used in this block is: Add Sub

Operation

Combo

Molt Elements Molt Elements Div Inv Sub Elements Inv Div

40

EyesWeb Compendium 2004

2.23 Math.Matrix.Concatenate Input/Output Datatype:

Description: This block combines 2 matrices in extension of each other row or column-wise. TIP: This is e.g. often useful when wanting to combine 2 sets of x,y coordinates to a 2x2 matrix describing a rectangle.

Parameters: Name

Type

Description This param set the block to work in one of the following modes:

Mode

Combo

Row column

41

EyesWeb Compendium 2004

2.24 Math.Matrix.DomianConv

Input/Output Datatype:

Description: This block converts the content of a Matrix from either ‘Real To Integer’ or from ‘Integer to Real’ numbers. TIP: This is e.g. used in situations where only one input type is accepted by another block.

Parameters: Name

Type

Description Type(0) converts the input from integers to real numbers.

Type

Combo Type(1) converts the input from real numbers to integers.

42

EyesWeb Compendium 2004

2.25 Math.Matrix.Extract Input/Output Datatype:

Description: This block extracts a desired sub-matrix from an input matrix. This is done by setting the ‘top left’ and ‘bottom right’ entry numbers of the sub-matrix in the params of this block. TIP: Remember that the first entry of a 2d Matrix starts in [0,0] and NOT [1,1].

Parameters: Name

Type

Description

First_Col

Integer

The Col nr from which the SubMatrix should start.

First_Row

Integer

Second_Col

Integer

Second_Row

Integer

The Row nr from which the SubMatrix should start. The Col nr from which the SubMatrix should end. The Row nr from which the SubMatrix should end.

43

EyesWeb Compendium 2004

2.26 Math.Matrix.GetEntr Input/Output Datatype:

Description: This block extracts a single entry from an input matrix. TIP: Remember that the first entry of a 2d Matrix starts in [0,0] and NOT [1,1].

Parameters: Name

Type

Description

Row

Integer

The Row nr of the entry.

Col

Integer

The Col nr of the entry.

44

EyesWeb Compendium 2004

2.27 Math.Matrix.Input.Generator Input/Output Datatype:

Description: This block generates a matrix whose entries may be specified by means of the Matrix editor provided in the custom dialog. The number of columns is limited to 545. The number of rows is only limited by the available memory. TIP: Use this block if you need a matrix of a certain size as input for e.g. the Imaging.BlobAnalysis.TrackRegions block.

Parameters: Name

Type

Description

NumRows

Integer

The row size of the generated matrix.

NumColumns

Integer

The col size of the generated matrix. Type of the matrix entries:

Type

Type(0) Real

Combo

Type(1) Integer

Period(ms)

Integer

The period of the interval in which a matrix is generated.

45

EyesWeb Compendium 2004

2.28 Math.Scalar.BinaryOp Input/Output Datatype:

Description: This block executes a binary operation on the 2 scalar inputs producing a scalar output.

Parameters: Name

Type

Description The different operations that can be used in this block is: Add Sub Molt Div

Operation

Combo

InvSub InvDiv Pow Inv Pow Max Min Mod Inv Mod

46

EyesWeb Compendium 2004

2.29 Math.Scalar.BinaryOpLogical Input/Output Datatype:

Description: This block performs a binary logical operation on the 2 input scalars. The resulting scalar is of the same type as the input scalar. A condition is checked for each entry of the inputs. The corresponding output entry is zero if the condition is false, otherwise its value is different from zero. TIP: This block can be used in an ‘If sentence’ like we know it from programming.

Parameters: Name

Type

Description The different operations that can be used in this block is: == (equals) != (not)

Operation

Combo

> (larger) >= (larger or equal) < (less) <= (less or equal) Min Mod Inv Mod

47

EyesWeb Compendium 2004 Min Mod Inv Mod

48

EyesWeb Compendium 2004

2.30 Math.Scalar.RunningOp Input/Output Datatype:

Description: This block calculates a running operation on the last N samples of the scalar input stream. TIP: Use this block when doing statistic calculations

Parameters: Name

Type

N

Integer

Description Calculates the following running operations on the last N samples of the input stream. The different filter types that can be used in this block is: Sum

Operation

Combo

Sample Mean Sample Standard Deviation Sample Variance RMS Duty Cycle

Reset

Command

Resets the block to 0.

49

EyesWeb Compendium 2004

2.31 Math.Scalar.UnaryOpLogical Input/Output Datatype:

Description: Perform a logical operation on the input scalar value against the parameter 'Value'. The output is either 0 or 1 representing “true” or “false”.

Parameters: Name

Type

Description The different operations that can be used in this block is: == !=

Operation

Combo

> >= < <= Nop

Value

Double

The value to compare to.

50

EyesWeb Compendium 2004

2.32 Math.Scalar.ConstOp Input/Output Datatype:

Description: Executes the selected operation on the scalar input with a constant. TIP: use this block if you e.g. need to multiply a scalar flow with a constant or do any other constant action to a scalar.

Parameters: Name

Type

ConstValue

Double

Description The selected operation will be performed with this constant. The different operations that can be used in this block is: Add Sub

Operation

Combo

Molt Div Inv Sub Inv Div Pow

51

EyesWeb Compendium 2004 Inv Pow Max Min Mod

52

EyesWeb Compendium 2004

2.33 Math.Scalar.ThresCrossing Input/Output Datatype:

Description: This block outputs the input value every time it crosses (upward, downward or both) a threshold value. TIP: This block is very helpful if you what to activate something if a scalar flow e.g. rises above a certain point.

Parameters: Name

Type

Description The modes of the block are: Both

CrossDirection

Combo Upward Downward

Threshold

Integer

The value of the checked threshold

53

EyesWeb Compendium 2004

2.34 MoCap.Item.Extract Input/Output Datatype:

Description: This block extracts an item from the input MoCap list. It is typically used in blob tracking. The value output is a 1x4 matrix containing [x, y, width, height] that can be used to calculate the bounding rectangle of a blob.

Parameters:

54

EyesWeb Compendium 2004 Name Extraction_Mo de

Type

Description

Combo

ByLabel, ByIndex

Label

String

Index

Integer

EnableList

Boolean

Labeled matrix to extract, if bylabel mode is selected If ByIndex is chosen, the index chooses the wanted matrix.

55

EyesWeb Compendium 2004

2.35 MotionAnalysis.Cues.SMI Input/Output Datatype:

Description: The SMI block receives an image flow, if there is motion in the flow it gives the output between 0,01 and 1, if there is no motion it gives the output 0. This can be used in e.g. motion detection.

Parameters: Name

Type

Images

Integer

Description The number of frames the block uses to calculate the current Silhouette Motion Image. The parameters for a median filter applied to the

Median Filter

Integer

calculated Silhouette Motion Image in order to reduce noise.

56

Related Documents

Eyesweb Compendium
October 2019 27
Telecom Compendium
December 2019 20
Compendium | 1827
June 2020 13
Colorado Compendium
May 2020 17
2 Compendium
May 2020 15
Compendium Daemonii
November 2019 23