bike theft news

On Wed, 21 Jun 2017 09:59:16 -0000 (UTC), Duane
wrote:

You can have glass break sensors turn the lights on.

Yep, that would work nicely. I had a paragraph written on alternative
methods and their complications, but decided to remove it to keep my
rant from becoming too complicated.

In most cases, a sensor would be redundant because the regular burglar
alarm would also have a glass breakage sensor that turns on some
lights, so that the police don't need to grope in the dark trying to
find where in the store the burglar is hiding, and so that the store
with the break-in is easier to identify from the street. However,
using a glass breakage sensor to turn on some lights does not do
anything for the long time it takes for a DVD to switch from b&w low
light, to bright store lights.

Notice that I said "some" lights. The problem is that most stores do
not have a centralized method of turning on *ALL* the lights. When
the alarm is triggered, it usually turns on one small light, often in
the wrong location. To make this idea work, all the lights would need
to be centrally controlled, which is possible, but currently uncommon.


--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

On 21/06/2017 7:57 AM, Jeff Liebermann wrote:
On Wed, 21 Jun 2017 09:59:16 -0000 (UTC), Duane
wrote:

You can have glass break sensors turn the lights on.

Yep, that would work nicely. I had a paragraph written on alternative
methods and their complications, but decided to remove it to keep my
rant from becoming too complicated.

In most cases, a sensor would be redundant because the regular burglar
alarm would also have a glass breakage sensor that turns on some
lights, so that the police don't need to grope in the dark trying to
find where in the store the burglar is hiding, and so that the store
with the break-in is easier to identify from the street. However,
using a glass breakage sensor to turn on some lights does not do
anything for the long time it takes for a DVD to switch from b&w low
light, to bright store lights.

Notice that I said "some" lights. The problem is that most stores do
not have a centralized method of turning on *ALL* the lights. When
the alarm is triggered, it usually turns on one small light, often in
the wrong location. To make this idea work, all the lights would need
to be centrally controlled, which is possible, but currently uncommon.



I've done security systems where this is the case. You don't actually
need all of the store lights on when that isn't practical. You can set
up some strategically placed flood lights. Certainly not brain surgery
to set up lighting to cover a camera's scope.

On 6/21/2017 7:42 AM, Jeff Liebermann wrote:
On Wed, 21 Jun 2017 00:11:12 -0400, Frank Krygowski
wrote:

On 6/20/2017 10:49 PM, Jeff Liebermann wrote:
On Tue, 20 Jun 2017 11:27:27 -0400, Frank Krygowski
wrote:

Motion-activated lights indoors would make that surveillance camera a
lot more useful!

Not really. The problem with aiming a motion activated surveillance
camera through a window is that everyone that passes by sets off the
motion detector.

What I had in mind was a motion detector controlling a light, not
(necessarily) controlling the camera.

Ummm... if the shop is dark, how is the camera going to see the
burglar so that the motion detector will work?

I'll try again:

My side porch light is controlled by a motion detector. If it's dark
outside and we walk up to the porch, the porch light comes on automatically.

I've also got a light in my basement workshop that's controlled by a
motion detector. If I walk into my dark workshop carrying something
heavy in both arms, I don't have to flick the light switch. The light
comes on automatically.

I'm proposing putting that system into the bike shop, controlling a
floodlight pointed the same direction as the camera. That way, the
camera should be filming an illuminated subject, not one in shadows. It
costs about $20 to try it.

--
- Frank Krygowski

On 2017-06-21 08:15, Frank Krygowski wrote:
On 6/21/2017 7:42 AM, Jeff Liebermann wrote:
On Wed, 21 Jun 2017 00:11:12 -0400, Frank Krygowski
wrote:

On 6/20/2017 10:49 PM, Jeff Liebermann wrote:
On Tue, 20 Jun 2017 11:27:27 -0400, Frank Krygowski
wrote:

Motion-activated lights indoors would make that surveillance camera a
lot more useful!

Not really. The problem with aiming a motion activated surveillance
camera through a window is that everyone that passes by sets off the
motion detector.


Very easily solvable.



What I had in mind was a motion detector controlling a light, not
(necessarily) controlling the camera.

Ummm... if the shop is dark, how is the camera going to see the
burglar so that the motion detector will work?


- LED lighting.


I'll try again:

My side porch light is controlled by a motion detector. If it's dark
outside and we walk up to the porch, the porch light comes on
automatically.

I've also got a light in my basement workshop that's controlled by a
motion detector. If I walk into my dark workshop carrying something
heavy in both arms, I don't have to flick the light switch. The light
comes on automatically.

I'm proposing putting that system into the bike shop, controlling a
floodlight pointed the same direction as the camera. That way, the
camera should be filming an illuminated subject, not one in shadows. It
costs about $20 to try it.


That can be annoying because it can trip every time someone comes
walking by too close to that window. There are much better alternatives
such as glass break sensors. Or an invisible IR beam like garage doors have.

The solutions here are simple. It just seems that the folks in the
industry serving this market aren't always the smartest ones. All most
do is market same-old same-old.

--
Regards, Joerg

http://www.analogconsultants.com/

On Tuesday, June 20, 2017 at 7:49:12 PM UTC-7, Jeff Liebermann wrote:
On Tue, 20 Jun 2017 11:27:27 -0400, Frank Krygowski
wrote:

Motion-activated lights indoors would make that surveillance camera a
lot more useful!

Not really. The problem with aiming a motion activated surveillance
camera through a window is that everyone that passes by sets off the
motion detector. The better DVR (digital video recorders) have
programmable zones. The viewing area is chopped into small
rectangular zones, each of which can be set to detect motion or ignore
motion. This helps, but is far from perfect. Something like this:
http://surveillance.aver.com/Images/Shared/IMD%20revised.jpg
where only the door areas detect motion.

The big problem with camera placement is the field of view usually
ends up many times wider than a person's head. My guess(tm) is the
camera field of view at the window was probably 20 times the width of
a person's head. If the camera was a better quality 1920x1080, the
persons head would be only:
1920 / 20 = 96
pixels wide. Try to identify someone from a photo where the person's
face is only 96 pixels across. Here's me at 660 pixels across:
http://802.11junk.com/jeffl/pics/jeffl/slides/jeffl-07.html
and at 96 pixels across, expanded to the same image size:
http://802.11junk.com/jeffl/pics/jeffl/jeffl-07-96.jpg
96 pixels across actually looks fairly good, but only under ideal
conditions. Add to that the distortion caused by positioning the
camera at about 45 degrees downtilt near the ceiling, marginal night
time lighting, and a possible disguise.

Also, the same camera would easily produce a recognizable face photo
if the field of view was reduced to perhaps 5 times the width of a
person's head, but then the field of view at the front window would be
too small to be useful and require multiple cameras to cover the
entire window.

Incidentally, my favorite mistake is to put cameras on the ceiling,
which usually produce great a video of the top of the burglars head,
hat, or hoodie.

There's also the problem that often in order get longer recording time they will reduce the dots/inch.

On 2017-06-21 13:30, wrote:
On Tuesday, June 20, 2017 at 7:49:12 PM UTC-7, Jeff Liebermann
wrote:
On Tue, 20 Jun 2017 11:27:27 -0400, Frank Krygowski
wrote:

Motion-activated lights indoors would make that surveillance
camera a lot more useful!

Not really. The problem with aiming a motion activated
surveillance camera through a window is that everyone that passes
by sets off the motion detector. The better DVR (digital video
recorders) have programmable zones. The viewing area is chopped
into small rectangular zones, each of which can be set to detect
motion or ignore motion. This helps, but is far from perfect.
Something like this:
http://surveillance.aver.com/Images/Shared/IMD%20revised.jpg
where only the door areas detect motion.

The big problem with camera placement is the field of view usually
ends up many times wider than a person's head. My guess(tm) is
the camera field of view at the window was probably 20 times the
width of a person's head. If the camera was a better quality
1920x1080, the persons head would be only: 1920 / 20 = 96 pixels
wide. Try to identify someone from a photo where the person's face
is only 96 pixels across. Here's me at 660 pixels across:
http://802.11junk.com/jeffl/pics/jeffl/slides/jeffl-07.html and
at 96 pixels across, expanded to the same image size:
http://802.11junk.com/jeffl/pics/jeffl/jeffl-07-96.jpg 96 pixels
across actually looks fairly good, but only under ideal conditions.
Add to that the distortion caused by positioning the camera at
about 45 degrees downtilt near the ceiling, marginal night time
lighting, and a possible disguise.

Also, the same camera would easily produce a recognizable face
photo if the field of view was reduced to perhaps 5 times the width
of a person's head, but then the field of view at the front window
would be too small to be useful and require multiple cameras to
cover the entire window.

Incidentally, my favorite mistake is to put cameras on the
ceiling, which usually produce great a video of the top of the
burglars head, hat, or hoodie.

There's also the problem that often in order get longer recording
time they will reduce the dots/inch.


Why? You just record round-robin and when the glass brake sensor
triggers you make sure that things are no longer over-written _and_ are
immediately being transferred offsite via Internet. In case the perps
find the stack of DVRs and rip them out.

In this day and age you could record a whole week in hi-def. Our DVR in
the living room can record over 100 full-length movies in hi-def. Enough
for watching Westerns all week with zero sleep.

--
Regards, Joerg

http://www.analogconsultants.com/

On Thursday, June 22, 2017 at 12:37:13 PM UTC-7, Joerg wrote:
On 2017-06-21 13:30, wrote:
On Tuesday, June 20, 2017 at 7:49:12 PM UTC-7, Jeff Liebermann
wrote:
On Tue, 20 Jun 2017 11:27:27 -0400, Frank Krygowski
wrote:

Motion-activated lights indoors would make that surveillance
camera a lot more useful!

Not really. The problem with aiming a motion activated
surveillance camera through a window is that everyone that passes
by sets off the motion detector. The better DVR (digital video
recorders) have programmable zones. The viewing area is chopped
into small rectangular zones, each of which can be set to detect
motion or ignore motion. This helps, but is far from perfect.
Something like this:
http://surveillance.aver.com/Images/Shared/IMD%20revised.jpg
where only the door areas detect motion.

The big problem with camera placement is the field of view usually
ends up many times wider than a person's head. My guess(tm) is
the camera field of view at the window was probably 20 times the
width of a person's head. If the camera was a better quality
1920x1080, the persons head would be only: 1920 / 20 = 96 pixels
wide. Try to identify someone from a photo where the person's face
is only 96 pixels across. Here's me at 660 pixels across:
http://802.11junk.com/jeffl/pics/jeffl/slides/jeffl-07.html and
at 96 pixels across, expanded to the same image size:
http://802.11junk.com/jeffl/pics/jeffl/jeffl-07-96.jpg 96 pixels
across actually looks fairly good, but only under ideal conditions.
Add to that the distortion caused by positioning the camera at
about 45 degrees downtilt near the ceiling, marginal night time
lighting, and a possible disguise.

Also, the same camera would easily produce a recognizable face
photo if the field of view was reduced to perhaps 5 times the width
of a person's head, but then the field of view at the front window
would be too small to be useful and require multiple cameras to
cover the entire window.

Incidentally, my favorite mistake is to put cameras on the
ceiling, which usually produce great a video of the top of the
burglars head, hat, or hoodie.

There's also the problem that often in order get longer recording
time they will reduce the dots/inch.


Why? You just record round-robin and when the glass brake sensor
triggers you make sure that things are no longer over-written _and_ are
immediately being transferred offsite via Internet. In case the perps
find the stack of DVRs and rip them out.

In this day and age you could record a whole week in hi-def. Our DVR in
the living room can record over 100 full-length movies in hi-def. Enough
for watching Westerns all week with zero sleep.

Firstly - most of these burglar alarm cameras are totally enclosed and even usually wireless. So they record internally on what would be a thumb drive I suppose.

In order to get the longest period before over-writing occurs they will both reduce the frames per second and the definition of the picture. Remember that transmitting these pictures (when you may have 5 or 10 cameras) is resource limited and you want as much self containment as possible.

Battery life is also a consideration.

On 2017-06-22 13:41, wrote:
On Thursday, June 22, 2017 at 12:37:13 PM UTC-7, Joerg wrote:
On 2017-06-21 13:30, wrote:
On Tuesday, June 20, 2017 at 7:49:12 PM UTC-7, Jeff Liebermann
wrote:
On Tue, 20 Jun 2017 11:27:27 -0400, Frank Krygowski
wrote:

Motion-activated lights indoors would make that surveillance
camera a lot more useful!

Not really. The problem with aiming a motion activated
surveillance camera through a window is that everyone that
passes by sets off the motion detector. The better DVR
(digital video recorders) have programmable zones. The viewing
area is chopped into small rectangular zones, each of which can
be set to detect motion or ignore motion. This helps, but is
far from perfect. Something like this:
http://surveillance.aver.com/Images/Shared/IMD%20revised.jpg
where only the door areas detect motion.

The big problem with camera placement is the field of view
usually ends up many times wider than a person's head. My
guess(tm) is the camera field of view at the window was
probably 20 times the width of a person's head. If the camera
was a better quality 1920x1080, the persons head would be only:
1920 / 20 = 96 pixels wide. Try to identify someone from a
photo where the person's face is only 96 pixels across. Here's
me at 660 pixels across:
http://802.11junk.com/jeffl/pics/jeffl/slides/jeffl-07.html
and at 96 pixels across, expanded to the same image size:
http://802.11junk.com/jeffl/pics/jeffl/jeffl-07-96.jpg 96
pixels across actually looks fairly good, but only under ideal
conditions. Add to that the distortion caused by positioning
the camera at about 45 degrees downtilt near the ceiling,
marginal night time lighting, and a possible disguise.

Also, the same camera would easily produce a recognizable face
photo if the field of view was reduced to perhaps 5 times the
width of a person's head, but then the field of view at the
front window would be too small to be useful and require
multiple cameras to cover the entire window.

Incidentally, my favorite mistake is to put cameras on the
ceiling, which usually produce great a video of the top of the
burglars head, hat, or hoodie.

There's also the problem that often in order get longer
recording time they will reduce the dots/inch.


Why? You just record round-robin and when the glass brake sensor
triggers you make sure that things are no longer over-written _and_
are immediately being transferred offsite via Internet. In case the
perps find the stack of DVRs and rip them out.

In this day and age you could record a whole week in hi-def. Our
DVR in the living room can record over 100 full-length movies in
hi-def. Enough for watching Westerns all week with zero sleep.

Firstly - most of these burglar alarm cameras are totally enclosed
and even usually wireless. So they record internally on what would be
a thumb drive I suppose.


That would be the first big mistake of the alarm company or store owner.
All it takes is one serious whack with a basseball bat or a well aimed
shot from the 22 and the evidence is toast.


In order to get the longest period before over-writing occurs they
will both reduce the frames per second and the definition of the
picture. Remember that transmitting these pictures (when you may have
5 or 10 cameras) is resource limited and you want as much self
containment as possible.

Battery life is also a consideration.


Wireless? Battery? That sounds like one of those $499 specials with half
a dozen cameras and lots of marketing fluff. A proper system must be
installed with wires and functionally tied into other sensors such as
the ones for breaking glass.

Self-containment is not always a good thing for surveillance. You don't
want to enable the perp to destroy evidence, either by brute force or by
setting the store on fire. At that point the evidence must already be
stored securely offsite or at least at a safe place in the building.
That sometimes even catches those that are supposed to protect us. They
thought they had ripped all the surveillance and recording equipment out
but missed some:

http://ktla.com/2015/06/11/video-of-...attorney-says/

--
Regards, Joerg

http://www.analogconsultants.com/