November 16, 2006

Histogram.h done

The C++ histogram is done, complete with scaling dilution functionality :P

As yet pretty untested, but it compiles with no problems.


Header Files

Pixel and Colour are two header files that we have created and can be downloaded here.



the List header is a file that is located in


you will need to add an extra argument to your compile instruction

I believe it is along the lines of
-l C:\Dev-C+\Include\G+

I just got annoyed and copied the contents of C:\Dev-C+\Include\G+
into the C:\Dev-C++\Include


Just cleaning stuff up then will put in all in a nice rar file for all.


Here’s the ColourMask class. All tested and works fine.



Here it is, EdgeDetector.cpp

It works! took about 11hrs but we must getting there. Hows everyone elses going?


Presentation run through

Nick is putting together the presentation material and we’ve booked out Presentation Room B at university house on Monday 20th from 1pm till 3pm so that we can familiarise ourselves with the material and give it a couple of run throughs.

November 15, 2006

What everybody is doing now

Follow-up to Delegation of Work from DCS Robot Vision System

Who’s doing what has changed a bit so just so everyone knows:

ImageProcessor Tom
Histogram Chris
ColourMask – Hoi
ContourProcessor – Ant’s pretty much done this
EdgeDetector – Matt

Comment if any changes to this. Also lets collectively pat ourselves on the back for Tom’s previous post + comments which for a while was the most popular blog entry on warwick blogs…. nice…



Hi All,

Here is the ContourProcessor.cpp

It compiles and to my knowledge it functions correctly althought a large majority hasn’t been tested but its time for bed.

Hope everyone elses crash course in C++ and java to C++ conversion is going ok


November 13, 2006

Delegation of Work

Follow-up to C++ code so far from DCS Robot Vision System

OK guys, after a lot of deliberating, i think that the best way for us to split up the work is for each of us to convert one of the following classes into c++:

ImageProcessor – Me
Histogram – Hoi (so you can also add a histogram for the contourprocessor)

Remaining people if you could comment on this entry with which you will be doing, then we wont end up doing the same thing twice!

In regards to the formatting of image throughout processing, it will be easiest to stick with CImg, and we still need to decide how we will get around the RGB colour issue with the CImg library, as I can’t seem to get it to produce the RGB values of each pixel, and have so far been doing it based on percentages of colour in each pixel.
Let us all know if one of you finds a way around this. Otherwise for all methods assume the same types for parameters and returns, and as i said use CImg rather than NCCTable.

Any questions just post them up!!!


Question about code

Which parts of the java do we NOT need to convert to c++?

Is it just the FrameAccess and JpegImagesToMovie classes?

This would leave:

which definitely need converting, and depending on how we are going to store the image and colour values while we are processing, we may need equivalent classes to:

Just let me know your thoughts….


C++ code so far

Here is the C++ code so far.


The part which I have done is in normalize.cpp, where I have basically added to Ants code.

I will sort out what needs to be done before lunch today, and post on here or email you guys with the equivalent java method, and spec for the c++ method which needs to be created.

Hopefully we can get all this done by Wednesday, then we can start testing it on the Robot by the end of the week so we can have some preparation time for the presentation over the weekend.

October 2023

Mo Tu We Th Fr Sa Su
Sep |  Today  |
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31               

Search this blog



Most recent comments

  • The reflective writing is submitted individually so make sure you submit it yourself tomorrow mornin… by on this entry
  • Yeh Ant's charts look great. Not sure – Ant reckons 1650 max (10% over), so if anyone has any sugges… by on this entry
  • Wow, the report looks really good :) Do you think it will matter too much if it is a couple of hundr… by on this entry
  • nice, overs better than under ;) by on this entry
  • I'll do the other bits then: introduction/overview of person tracking, testing/problems (if Ant+Hoi … by Matt on this entry

Blog archive

RSS2.0 Atom
Not signed in
Sign in

Powered by BlogBuilder