Wednesday 31st August 2005

Scamera evidence is now online

Photos to go live on Wilts SCP Web site

---

Instant evidence now online

You can now see the photographic evidence against you if you are snapped breaking the speed limit. That's as long as you were in Wiltshire at the time.

Wiltshire and Swindon Safety Camera Partnership (WSSCP) said that it will send those accused of breaking the legal limit or running red lights details of its Web site and an individual PIN. This then allows those accused to view the evidence against them securely.

The site includes photographs maps of where alleged offences took place too. It's the UK's first site to provide online evidence.

• www.safetycameraswiltshire.co.uk

[turbobungle]

All scamera operators should have to do this as well as supplying details of the officer who nicked you and the serial no. of the equipment and the calibration details of the equipment with proof of reliability (which is very unlikely!)

[antispeed]

With the recent news about the encription code used to secure scamcans picture and video evidence is not as secure as the government would have us believe, How long before a standard plea of " not guilty because you believe the evidence has been tampered with " ?

ahhh silly me !!!! they will change the law to suit wont they?

[streaky]

antispeed said:

With the recent news about the encription code used to secure scamcans picture and video evidence is not as secure as the government would have us believe, How long before a standard plea of " not guilty because you believe the evidence has been tampered with " ? ahhh silly me !!!! they will change the law to suit wont they?

The 'issue' of 'collisions' in MD5 (and other hashing algorithms) has been discussed elsewhere. It is unlikely that any claim that the limited number of 'laboratory-engineered collisions' shown to exist would cast "reasonable doubt" on the evidence. No-one has shown that 'collisions' exist other than in carefully engineered constructs hashed using MD5 (and similar) - Streaky

[chrisjl]

streaky said:

The 'issue' of 'collisions' in MD5 (and other hashing algorithms) has been discussed elsewhere. It is unlikely that any claim that the limited number of 'laboratory-engineered collisions' shown to exist would cast "reasonable doubt" on the evidence. No-one has shown that 'collisions' exist other than in carefully engineered constructs hashed using MD5 (and similar) - Streaky

I didn't entirely understand that, which may affect the relevance of what follows, but...
Hasn't it been proved that a file can be deliberately constructed to have a desired MD5, invalidating the usual "the MD5 matches the file, so it can't possibly have been tampered with" check? Having that constructed file also be a valid JPG/PNG/whatever image of your car is an added complexity that may take this trick beyond the abilities of 99.99999%

[s a m]

I want to know how much they paid for that website.

Most of the scamera partnerships have got decent government looking sites, that one looks like its done by the local school kids.

[x1m]

Having been to school in Swindon for 5 years, it wouldn't suprise if it was done by local school kids. Getting involved with the community and what not.

[deadlym]

s a m said:

I want to know how much they paid for that website. Most of the scamera partnerships have got decent government looking sites, that one looks like its done by the local school kids.

Heh, yes - my eyes! Too much yellow!

www.safetycameraswiltshire.co.uk said:

DRIVE SAFE. DRIVE WITHIN THE SPEED LIMIT.

So that's all I need to do to drive safe(ly). No need to pay attention. <sigh>

[chimyellow]

Wiltshire Web Site said:

Moreover, in particular circumstances, driving at speeds lower than the legal limit may result in prosecution for other offences, for example dangerous driving or driving without due care and attention, when the speed is inappropriate and inherently unsafe.

Does this mean that they are going to introduce non-speed camera to catch all the people that drive at 45mph on a clear straight NSL road?

[zumbruk]

chrisjl said:

Hasn't it been proved that a file can be deliberately constructed to have a desired MD5,

No.

What has happened is that the authors of http://eprint.iacr.org/2004/199 have shown two different datastreams that produce the same hash; the datastreams are shown here; www.rtfm.com/movabletype/archives/2004_08.html#001055

Of course, because of the "pigeon hole principle" (see http://mathworld.wolfram.com/DirichletsBoxPrinciple.html) this was inevitable.

No-one has produced a general purpose crack of MD5 such as you suggest.

[james_j]

deadlym said:

s a m said: I want to know how much they paid for that website. Most of the scamera partnerships have got decent government looking sites, that one looks like its done by the local school kids. Heh, yes - my eyes! Too much yellow! <!--[qte1=www.safetycameraswiltshire.co.uk]--><blockquote><font size="1">www.safetycameraswiltshire.co.uk said:</font><table class="categoryCell" border="0" cellpadding="8" cellspacing="2"><tr><td class="altForumCell"><!--[qte2]-->DRIVE SAFE. DRIVE WITHIN THE SPEED LIMIT.

So that's all I need to do to drive safe(ly). No need to pay attention. <sigh>[/quote]

They really are ar$eholes aren't they?

[Andy_L]

turbobungle said:

All scamera operators should have to do this as well as supplying details of the officer who nicked you and the serial no. of the equipment and the calibration details of the equipment with proof of reliability (which is very unlikely!)

Calibration certificates are not enough all these do is state that at the time of the annual check the equipment was working within tolerances. It does not show whether or not the equipment has been modified since calibration and still in 'type approval' or whether it has a history of intermitant failure. The only way to be able to answer questions such as these is to see the service history or maintenance records for the equipment concerned - assuming that they have been kept up to date that is.

[kevinday]

Also bear in mind that most 'calibration' is not actually calibration, merely a test of the the distance measurement function (static test). The time elapsed function which is needed to calculate the speed is not calibrated at all.