Quote of the week

Advertisements

Why I moved away from FOSS as my main toolset

The road from Free and Open Source Software to a single, fit for purpose solution.

When I first started out investigating video within the police service I had very little knowledge and understanding of some of the complexities involved.

The requirement to learn was born out of necessity.

How could I get a better image from an old VHS tape and get it disseminated to other officers quicker?

I had no method to digitize the footage so, in order to figure this out, I started the great internet hunt. In 2003, that was a challenge in itself!

I had no budget, and literally had to beg, borrow and ‘acquire’ equipment. The first computer was one of my old ones from home!

I would read somewhere of how to do something, and then I would research what was required. This led me to the wonderful world of Free and Open Source Software (FOSS), and further down the road to the delights of the Linux operating system.

Spreadys.com was built on posts detailing my adventures with this type of software and how to complete certain tasks with them.

I built Spreadys Software Pack… and I still see this in use on people’s computers today!

20161027_164103

I trained people in various packages such as Scalpel, FFmpeg, Mencoder and Virtualdub.

Later, I showed how people can create their own FVA Toolbar, with all their little software packages and applications.

I have this toolbar – Here is the one on my desktop!

FVA-toolbar

It is a quick way to reach some of the many tools in my toolbox. Some of the tools are designed with a single function, some more powerful with multiple uses. All can be great fun to use and learn.

BUT, I don’t use them every day…I may not use them for a few weeks. However, I still, on occasion like to check something and need to utilize one or a few of them.

The move away from FOSS was again born out of necessity.

As an example, let’s look at the classic Non-Linear-Editor (NLE), used for cutting and editing video.

I started with Windows Movie Maker! I can hear the laughter now!!!

It was free, it came with XP (and previously with Windows ME – where I started with it), and it could do some of the things I wanted… although not very well!

What did I do when it didn’t do something?  – I searched and found other small programs… either free or very cheap.

It was not long before this process of using 4-5 different programs started to cause problems. I need one program that was easy to learn, and easy to train others in…. This was a key requirement as our video unit was growing.

Consequently, we received funding for some professional NLE licences.

I no longer needed multiple small applications and move from one to the other – I could do all, and more, from within a single application.

It was much easier also to train others…as the single application had a much easier learning curve.

We can now Fast Forward >>>>> Circa 2012/13

I had built a good knowledge base around various applications and even started a little batch scripting, (my level of programming knowledge!!!), to link various Command Line Interfaces together.

But… things were getting complicated, and in more ways than encountered during the NLE example.

  • Identifying the correct application to use, and the version number was becoming tricky. Research, testing and validation was required, especially on the applications that were getting updated a lot. Version 7.2 would work, but 7.3 would not…but that may work with something else!
  • Many applications utilized other components that were developed by 3rd parties. Finding the correct match would ensure a successful processing requirement, but it was always a case of trial and error….and it was up to me to do the testing.
  • Reporting on the use of multiple applications was very time consuming. Screen grabs, text files… any method to retain what settings were used. More difficult was the validation process to ensure that what was output from one application, was read correctly by another. I had many examples where XXX frames were output, but only XX frames were read by another!
  • Half of my time was being spent on research, testing, process-flow…rather than actually doing! Yes…. I enjoyed the challenge, and the results, but ultimately it affected how much I could get done.
  • Training and teaching others in the applications and methods was the next challenge. There are just so many variables when using multiple packages that it was impossible to account for all circumstances.
  • Validation of each application, each plug-in, each version and each configuration were extremely difficult and had to be done on a case by case basis due to the different variables involved.
  • I was the support. I would often spend hours researching blogs and forums to identify issues and solve problems.
  • Identifying problems was hard, and even more time consuming. Missing frames, missing data, corrupted frames, changes in colour and light. What was the software doing to my evidence?
  • I needed more. I had over 80 applications in my FOSS collection. Ranging from a small utility to hash a file, to an application to stack astronomy images for frame integration. However, I needed to do more, faster… and with evidential integrity.

I had built an extensive knowledge and toolset – but I needed a single solution.

multitool

I needed everything in one application, I needed the Swiss Army Knife, and the Leatherman multi-tool together… I also needed the support system. Importantly, I also needed it to be built for what I do – not another product made to fit but built fit for purpose.

I then discovered Amped FIVE, an application that had everything I needed and utilized some of the applications and libraries I was familiar with. It automated many processing tasks thereby removing user error and importantly, these processes were tested and controlled. The bonus was that it contained functions and enhancement powers that I never realized existed!

……. And the rest, as they say, is history.

What’s with all the screen snaps?

A screen snap is a digital image, taken using a digital camera, smartphone or bodycamera, of an other image being displayed on a computer screen!

I have talked about them before (here).

Over the past 6 months, there has been increase in these snaps appearing in media publications for the purpose of subject recognition.

The following are from January 3rd 2018:

https://www.yorkshirepost.co.uk/news/cctv-image-released-after-sainsbury-s-theft-in-ripon-1-8938247

1

The Moire effect is seen in the image, along with (possibly), color and detail loss. It also appears a little ‘squished’ – Perhaps the image was being displayed in a proprietary player and the video is distorted horizontally? It could be a simple Aspect Ratio issue.

Although many things can be restored and enhanced – It would have been easier to work on the original video export.

1a

Side by Side ‘Before-After’ in Amped FIVE

https://www.yorkshirepost.co.uk/news/cctv-appeal-after-new-year-s-eve-burglary-in-halifax-1-8938483

image2

Another Image with considerable processing damage..

2a

Split view of filter results in Amped FIVE

I would guess that both of these images came from video…. and as such there would have been more images to use. Even working on 2 or 3 gives a better chance to integrate, and bring out details lost in a simple frame export.

With FIVE’s Temporal Smoothing, Motion Smoothing and Frame Averaging filters – it’s only a few mouse clicks.

It is obvious to me, that this increase is partly caused by the decimation of dedicated CCTV and Video Units and the pressure on front line policing resources to do ‘something’ in order to get the subject recognized.

If you are working in a Media Unit – dealing with Video and Images from CCTV everyday, for the purpose of publication, remember that there is dedicated software to help you deal with them.

CCTV is often installed incorrectly, giving below par images to start with. Processing them incorrectly and making them worse seems a little counter productive!

..and for the videos that have come from correctly installed systems, just snapping the CCTV Monitor with your iPhone can negate all their hard work to provide quality evidence.

Not only that, but not working on the true digital export (bit for bit copy), could lead to questions further down the line concerning integrity and authenticity.

Many of us have seen news reports recently on the pressures of front line staff.

https://www.standard.co.uk/news/crime/police-impose-20-minute-limit-on-examining-cctv-for-petty-crimes-a3729811.html

I have often stated that dealing with CCTV should not be one of those pressures. They should be able to turn to the staff and officers who are dedicated to that task. I am all too aware that those people are easy targets for cost cutting bosses and those that are still doing the job are doing the work of three people… and still have to deal with the major incidents that come in through the door!

So, if you are one of those people, one of the few people left, working day-to-day on CCTV Videos and Images – let’s work towards presenting better images and video to the public, and in turn perhaps get a few more people recognized and all without risking integrity & authenticity questions.

Forensic Europe Expo 2016

Image 15

This week it is the Forensics Europe Expo at London’s Olympia.

I will be there with Amped Software and as such, it would be a fantastic time to pop by, catch up and see how the Amped Software products can assist your digital image and video investigations.

I will be presenting on Wednesday afternoon, “It is more than just a video”. This will cover the importance of original evidence, integrity and maintaining a methodical,scientific workflow.

No heckling or the throwing of heavy objects allowed!

Hope to see you sometime over the two days.

fee_logo16

 

 

 

Licence Plate Recovery

I’m always up for a challenge!

This tweet perked my interest yesterday..

Image 10c

 

‘Try it for yourself’ it said – OK then!

Now, you may be wondering why the VideoCleaner version is censored. I have today (27/3/2016) received notice that this number plate can no longer be shown publicly. As a result – you can no longer see the version above or the results below. 

What was even more interesting though was when I read the following at the VideoCleaner support Linkedin Page..

“VideoCleaner was the only program that could recover the license plate from this video”.

For full transparency – I have also now been informed that this wording was an error – It should have stated that it was, ‘the only Free and Open Source Program that could recover the licence plate’.

Now, as most readers of my blog posts here at spreadys.com will know, I am not adverse to free and open source software. I have been using them all for years. I have posted many articles on how to use X to get Y etc etc…

I am also now the International Trainer for Amped Software, The Forensic Image Processing Company.

I always conduct comparative work between Amped FIVE and other FOSS tools, not only to test image quality, but functionality, speed of use and evidential integrity.

So, the question immediately being asked was – Can FIVE do it?

Scrubbing through the 100 frames revealed static pixels under the generated noise.

Image 003

Highlighted area of static levels

By viewing the saturated pixels I could clearly see areas down the right side which did have values.

Image 005

Image 004

After those visual clues were identified it took 2 mouse clicks!

Image 006c

 

About a minute later – and a few more filters and formatting….

XVID-160324122610c

 

In conclusion – the question being asked was – could FIVE do it?

Answer – Yes!

…and that’s a good thing. Forensic Science is all about repeat-ability. You must be able to reproduce a conclusion if required. I have reproduced Doug’s licence plate. It’s subjective, but I prefer the visual presentation of FIVE’s result. The processing time from load, through analysis, all the way to output was only a couple of minutes. Most of that was analyzing the video to establish a filter chain!

Edited on 27/03/2016 to hide licence plate and explain reasons for censorship.