WestPoint · Cameras · Dallmeier
Dallmeier covers the whole apron with a single unit. If there is somewhere to put it.
German, and with an idea of its own: instead of spreading ten cameras over a large surface, one unit with several lenses of different distances that gives similar detail near and far. That changes the whole project, not just the purchase.
What kind of manufacturer it is.
Its approach is the multifocal sensor: several lenses with different focal lengths inside the same body, and a system that composes a single image out of them. What you get is detail that is similar at twenty metres and at a hundred and twenty, which no single camera does.
That attacks a real problem with any panoramic view: the more it takes in, the less detail is left far away. A very high resolution camera in the corner of an open compound sees the whole of it and does not let you recognise anybody at the far end. Here the detail is shared out by zones, and that sharing out is worked out and signed off before buying.
What it sacrifices is simplicity and price per unit. These are large, heavy, expensive units that ask for a serious bracket, real bandwidth and a workstation able to draw that image. The saving, when it appears, is not in the camera: it is in how many brackets, how many cable routes and how many fields of view have to be maintained afterwards.
What changes in the installation.
Everything depends on one thing: whether there is a point from which the whole surface can be seen. A mast, the corner of a building, the roof of an industrial unit. If that point does not exist, or if a column, a crane or a parked trailer covers a third of the scene half the day, the idea falls down and you go back to cameras spread out. That is decided on the visit, looking from up there and not from the plan.
And the calculation stops being a number on a sheet. It is done over the real surface, zone by zone, saying where a person has to be identified and where seeing that something is moving is enough. That document is the project, and on handover day it is checked with a person walking at the furthest point, which is the only test worth anything.
Where it behaves well.
On large open surfaces with a good vantage point: vehicle compounds, manoeuvring yards, terminals, quays, outdoor car parks, squares, sites with an occupancy figure. Places where the alternative is eight brackets with their eight cable routes and their eight fields of view to check every year.
Where we would not put it.
It is a tool for one particular problem, and outside it the cost is high.
In compartmented interiors —corridors, offices, store rooms, staircases— it makes no sense: what is needed there is small, cheap cameras in every gap, because the problem is not the distance, it is the walls.
Nor in small or medium installations where the number of cameras does not make up for the cost of a unit like this, or where there is no high, firm place from which everything can be seen. And we do not propose it when the budget does not include storage and a workstation to match: an image that size, badly recorded or badly drawn, is no use, and the camera gets the blame.
The analytics it carries inside, and where they stop.
A unit like this is good starting material for analysis: uniform detail, a single hour and a single scene, without the jumps that appear when you stitch together what eight different cameras see. Intrusion detection in zones and counting over that image are the natural use.
The limit is of two kinds. The usual one: what can be drawn is what gets solved. And the one of size: analysing an image like this costs machine power, so you have to decide which zones are analysed instead of asking for all of it at once.
What comes now is a recommendation with an interest of our own, so it is declared: in the same house there is IRIS Neural, the NVMS from Infinity Neural, the other company in the group. It is our own product. We would rather you knew it by reading this line.
What it brings is the understanding of the scene over that same image, and the decision about which zones deserve attention and which do not. And if what was needed was to cover the compound and be able to reconstruct what happened, the unit on its own already does that.
What we do.
- The visit to look for the vantage point, with the height measured and the obstructions noted down at the hours when there are lorries, cars and people. If there is no point, we say so and we change the approach.
- The sharing out of detail by zones over the plan of the surface, saying what will be possible in each one: identify, recognise or only see. That is handed over in writing before asking for a price.
- The bracket and the structure: a heavy unit on a mast that moves in the wind moves the field of view where it does most damage, which is far away. The rigidity is calculated, not estimated.
- The network, the recording and the workstation of whoever is going to look at that image, sized for real. And the access for maintenance: if a platform has to be hired to clean the glass, that is an annual cost and it goes in writing from the start.
- The maintenance: cleaning the glass, checking the field of view and the state of the structure.
The angle and the height are not decided at fitting time.
With a unit that covers more than a hundred metres, two degrees of tilt are a lot of metres gained or lost at the far end. And the height is a compromise: higher up covers less and looks down more steeply —worse for recognising a face—; lower down, the face is better and it is easier for a lorry to hide half the apron.
That is settled in the project, with the plan and the height measured, and not on the day of fitting. It decides for good what will be possible with that image, and it is the one thing in the installation that cannot be corrected afterwards without paying for the crane again.
Where to go next.
Before deciding how many cameras cover a place, you have to decide what detail is needed at every metre. That is what the area talks about.
Get started
Do you have an apron that never quite gets covered?
Tell us how much surface it is, where it could be looked at from and what you would have to be able to do with that image: recognise somebody, follow a vehicle or just know there is movement. With that we will tell you whether a unit like this makes sense here or whether four ordinary cameras come out better.