While a "dashcam" is nominally about visual images, newer automotive technology is looking at radar and other technologies in proximity sensors. Perhaps, particularly if the Control/storage is separated from the sensors,
While a "dashcam" is nominally about visual images, newer automotive technology is looking at radar and other technologies in proximity sensors. Perhaps, particularly if the Control/storage is separated from the sensors
Most of us want to be able to read the license plates, and on current 360 degree cameras you can not read any license plates!
A good front + rear camera will catch most accidents that are not your own fault and will catch most people who try to escape from a side impact even if they don't see the impact.
Once dashcams can write 200 Mbps to 1TB memory cards then it may be worth thinking about - another 4 or 5 years...
The SD UHS bus is already fast enough for 200mb/s, many cards can write 80 MBytes per second (640 mb/s). The problem is our dashcams still use the basic SD bus and don't use UHS. We need the dashcam processors updated to support fast UHS and 8K video so that there is enough resolution for 360 degree video. 4K video isn't enough to be able to read number plates easily on 360 degree images.
Or maybe it is better to use 5 separate cameras for front/rear/left/right/interior and some clever software on your phone/computer to combine them into a single 360 degree video? We should be able to do that now, but it would be much easier if we could have 4 small cameras like the mini0906 rear camera and 1 super wide angle like the Waylens for the interior, all connected to a central black box containing the 1TB microSD card.
We don't need to record the 8K vertical for a dashcam, 1080 is OK for vertical and 1920 * 4 = 7680 horizontally.
1080 * 7680 = 8,294,400 which is the same number of pixels as a normal 4K image and 100mb/s should be OK, so not far beyond today's technology. We just need a processor that can collect the 1080 frames from 4 remote 1080 cameras and join them side by side to make the 1080 x 7680 image.
Of course 1080 is only just good enough, what we really want is 2160 x 15,360 = 33,177,600 at 400mb/s (Same number of pixels as an 8K image.)
If you make a module independent of the two cameras, so that the cameras are smaller and discreet, you could integrate a hard disk into the module instead of a memory card and the gps module that was independent, so the capture of gps coverage would be almost instantaneous to be able to place the gps antenna away from metal and interference
My suggestions are:
1) - settings, viewing the current image - WIFI.
- viewing the history of the images - initially you can via - WiFi (but the quality of the viewing deteriorates sharply)
2) - for car radio - in addition: the mode of viewing the current image, controlling the settings, through otg with support for this function on the radio.
3) - copying files from the DVR to the USB flash drive (or external external device) via otg with the support of this function on the radio.
Here are my options for connecting the mini-0906 to the car receiver (for now, without Wifi and managing the settings): http://4pda.ru/forum/index.php?showtopic=876532&view=findpost&p=71909912
4) - for two or more cameras:
- or the picture mode in the picture (for 3 or more cameras: mode 2, 3 ... pictures in the main picture:
- either to build images based on the principle of a circular view system, each camera has its own places on the screen
5) - GPS - make an external connection. On car radio, many already have these antennas, and they are located in the right places. Why on one car have two GPS antennas?
6) For option 3 or more cameras:
- It is much more convenient to place the video signal processing unit not on the windshield, but under the torpedo or seat. Those. do it without a screen at all:
- on a smartphone or car receiver (paragraph 1 - item 4)
But this will already be another DVR (not mini-0906)
How about an Enterprise Dashcam?
Consumer dashcam is high volume/low price.
Enterprise can be high volume/high price.
But enterprise models have different requirements. Perhaps this is too radical redesign and too much for you to handle.
Main unit should be a separate blackbox (think "Raspberry Pi" or "Banana Pi" for the size) that can be hidden from view.
I think this alone could make you a market leader. Instead of a small main unit that is camera and CPU and GPS and display and front camera and ... ... ..., which becomes overheated due to too much workload, you can make a dedicated main unit inside a blackbox that is hidden in the car, for example inside glovebox or underneath center console.
For enterprise use, this kind of form factor can prevent many kinds of manipulation (removing of TF card to destroy evidence, disconnect cable to hide illegal activities, etc.) by bad workers.
This form factor also gives you new possibilities while still remaining "Mini". The cameras can be connected to the blackbox via cable, so front and rear only has a small camera and no other distraction. This is very discrete.
Also the blackbox form factor allows you to provide a SATA port for SSD instead of using TF card. There would never be an issue with broken TF card again. SSD can support much more writing than TF card and has higher performance.
The display with buttons and GPS can be optional add-on modules that are connected via cable for the user who needs them. This design allows you to build a wide ecosystem of compatible add-ons (high quality cam, low quality cam, 360 cam, IP67 outside cam, big display, small display, battery pack, etc.)
With this form factor, you can support more than two cameras. For example, you could provide 5 inputs. (FRONT, REAR, INTERIOR, LEFT, RIGHT). Then it could be useful to monitor large trucks (LEFT and RIGHT is useful to record evidence of crash with bicycles who are not seen by truck drivers in the mirrors). Users who don't need so many inputs just leave it disconnected.
Configuration should be by WiFi and small web server. WiFi password can act as security. Please don't make an app. Apple and Google change their OS all the time, and you have to update your app or it will stop working in a few years. This is the biggest problem with Chinese products at the moment: the vendor stops to care and a product stops to work with recent OS. If you make a web based configuration with mobile design, you never have to worry about upgrading the app. All smartphones have a good web browser and will always have it.
Unfortunately there is probably no SOC that can support this kind of ideas (I have many more). You would have to invest substantial R&D to make such a hardware design.
But I'm sure there is a market for this kind of product, maybe even possibilities of OEM for large truck companies if you have the right connections.
Here is a simpler suggestion for the next model:
Consider a smaller camera connection cable. Many vehicles in Europe have a kind of rubber tube to connect the car body to trunk door. This tube houses the cables to the rear lights, etc. It is the perfect place to do wiring of a rear camera, but with the current 0906 connection cables, it is difficult to put the wire through this tube. Perhaps use Mini SMB connector to decrease the size of the connector as much as possible.
1. I would like to add if you can make the dashcam work better in temperature conditions. Especially in high heat where dashcam overheats easily. Because here in Phoenix, Arizona it gets hot as 50C. My cheaper G1W can perform better and it's ridiculous.
2. The video quality is also terrible when you're driving against the sun, even with CPL filter. I drive to and from work and every time the sun is pointing at me , it makes very hard to read license plates of other vechicles even when I drive at a slow speed.
I would say stick to basics. My biggest beef with Mini series has been quality control. Build a camera that works as advertised, reliably and consistently. The moment I have to refocus the lens, change resistors or anything else to make it work is a non starter.