From what I have seen and learnt about thermal imaging cameras over the past few years of repairing them, at the base level, they are actually not that complicated. The clever stuff is usually the signal processing to achieve the best possible image from the collected data. If the camera is not for accurate radiometric use, its measurement accuracy becomes less critical and it may not even need a temperature read-out. Such versions are used for SOLAS and weapons sights where temperature differentiation is required and not spot temperature measurement.
The Microbolometter is the key to success here. If you can source a microbolometer with decent resolution and performance, at an affordable price, you have overcome the first significant hurdle. You then need a half decent lens. These used to be extremely expensive due to the Germanium that was used to create them. With the advent of new special glasses that may be moulded the price has dropped significantly (but they still aren't cheap). Microbolometers are a controlled product, n=both commercially and in terms of dual use technology regulations. That makes them pretty hard to source unless you have the appropriate credentials. They are also expensive when compared to a more common CCD chip.
Once you have the optical and detector block mounted you need to read-out the values from the microbolometer and process them. An ADC may be external or internal to the Sensor assembly dependant upon the part sourced.
The 'clever' stuff is in calibrating the microbolometer output and flattening the base line response using Non Uniformity Correction. A simple calibration shutter was used but more complex modern designs are able to work without such.
As Mike has shown, if you are able to read-out the data from the sensor, you can manipulate it as you wish. The LEPTON is an advanced design in that it has been highly integrated by FLIR and much of the hard stuff is done for the designer. What must be remembered here is that FLIR had the knowledge and money to design such a sensor package. Mu do not have such resources available to them and were so say using a commonly available sensor that had been on the market for some time. If such a sensor actually existed, it would be unlikely that it would be as friendly to use as the new LEPTON device.
I called the Mu project bogus when I saw the projected retail price as I could not see how the microbolometer of the specifications given, and the lens assembly, could be procured by them to make such feasible. At the time, no such combination was available at an affordable price for the Mu project. I stand by that view. Even the LEPTON with its high integration and new lens design is costing far more than Mu would have needed to meet their claims .... and it is only 80 x 60 pixels with 9fps.
I don't think anyone in the industry thought that a small company could not produce their own thermal camera, provided they were given access to the required controlled technology, it was the projected final selling price that appeared unrealistic.
It is my personal opinion that Mu should have produced a pretty reasonable thermal imaging product with the money and time that they had available to them. It is also my personal view that they did not have enough money to place such a design into large scale production in order to reduce the BoM coats and that the projected selling price of the product was wildly optimistic as a result.
A classic case of blue sky thinking escaping the concept phase without someone at the helm managing expectations.
And I still say that considering the time they have claimed to have spent on the case....its a damned ugly beast. They have absolutely no excuse for this as creating a custom case of good quality is not rocket science these days and the claims they have made on that topic have not been shown in reality. Much like their whole sorry project really.
Home
Help
Search
Login
Register
