This information HAS errors and is made available WITHOUT ANY WARRANTY OF ANY KIND and without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. It is not permissible to be read by anyone who has ever met a lawyer or attorney. Use is confined to Engineers with more than 370 course hours of engineering.
If you see an error contact:
+1(785) 841 3089
inform@xtronics.com
There is no way to know if someone else thought up these ideas before, but see the date on when they are added to this page in case attacked by patent vultures. I don't think I will ever get around to developing these ideas so putting them here might help someone else.
Brake lights tell the drivers of cars behind you that your foot is on the brake, but they don't tell how hard you are pushing. Such information could be coded into the brake light so that a crash avoidance system would have effective information. There could also be a display to let the following driver have some indication of how fast your car will slow down.
The idea requires one to imagine two places that need to have the data synchronized. A change is made in the first location - perhaps some code is changed. The first location now has two sets of the data. A file is generated that has these differences encoded by many different methods. By looking at the files that have changed deletions are rather simple - just start and end points of what is removed. What is more challenging is to add data.
The program has an extensible library of methods. One method of expansion can be to look at all the files on the system and see if the addition exists in any of them. A small starting block is located in any file common to both images and attempts to expand it are attempted. If a practical sized block is found, the file location and start and stop point to copy are put in the diff file with the insertion point and file to insert it into.
Blocks of data can also be manipulated before they are compared. An example would be adding an offset to he entire block. All offsets can be tried until the best performance is found. Blocks and also be manipulated by adding wrapable increasing numbers to the block, or geometric or pseudo random offsets. The blocks could also be reversed, as a first step. Most any math operation that transforms blocks of data can be used. The amount of CPU time would be extensive while searching for a match, but the payoff is distribution of the change would use very little bandwidth. The code to support methods of decompression can also be sent to update the receiving location.
Security fences, such as might be at a countries boarder, are difficult to monitor due to their great length. If they are constructed with horizontal elements that are conducting yet insulated or a simple insulated wire or even coaxial cables could be woven into the fence to form horizontal elements or horizontal hollow metal tube that have a single wire down the length, breaks and even the exact location of the break can be determined using 'time domain reflectometry'.
A square wave pulse can be sent down the wire or cable and the reflection can be detected even if the other end is on the order of 100 miles away. The time it takes for that reflection to return is dependent on the length of the wire. If that length changes, we know there is a break in the wire. The new length of time indicates with precision where that break is. Even if the intruder splices the wire after cutting it - the discontinuous impedance will create a smaller reflection indicating the point of the splice.
When I first had this idea the electronics to support it was problematic due to the speeds involved. Today it should work with out a problem. A laser beam is split into two one part reflects off of a diaphragm (can be most any type - ribbon dome etc. The reflected light and the other half of the LASER beam are combined and form an interference pattern that has rings that move in and out as the detection beam is modulated.
Two optical sensors are placed on a radius line spaced at a distance equal to 1/2 the space between the interference rings. The output of these sensors produce a quadrature wave form that drive an up/down counter of 16 bits (or more).
The worst case frequency out put of the quadratures at 20khz 16 bit full scale audio is 20k * 2 * 2^16 or 2,621,400,000 or about 2.6Ghz. This is a reasonable clocking speed and there are fiber optic receivers available to work at these speeds.
There needs to be a counter buffer so we don't read while the data changes. Reading this buffer at 40khz should give us audio that has absolutely zero preamp distortion!
A regulated DC power supply is connected to a PTC themistor with a temperture knee some where above ambient. The PTC works as a very stable an accurate temperature regulator - it gets to and maintains a constant temperature. The PTC should ideally be aluminum or gold plated to maintain a low Infrared emistivity. The wires to the themistor should be as thin a gage possible so as to prevent heat conduction. The thermistor should be mounted by only the supply wires to provide a high thermal resistance to the vacuum chamber.
Measuring the current will allowing calculating the amount of power needed to maintain temperature. As radiation is minimized, the only cooling is by conduction - which is dependent on gas pressure.
Twisting wire nuts can cause repetitive stress injury. If the closed end was shaped as a 1/4" hexagonal nut it could be driven with any old electric screw driver increasing the speed of installation and preventing injury.
You have all seen these, power adapters that plug into the AC mains that provide a bit of AC, DC (sometimes regulated) power for electronic items. The black blobs on the wall reminds us of warts ; thus wall-warts.
The improvement would use a switching power supply with a twist. When you plug it in the wall it starts operating just like any other wall wart - but if it sees no or a near zero load it shuts down for some period of time to save power. A twist of this would be it shutting down for increasing lengths of time to save even more power. The power for the timer can be derived by a high value resistor or small value capacitor coupling the part directly to the AC mains - using almost nothing compared to the several watts of wasted power in a standard
This idea was originally for microscope applications, but is clear
that it could be used for any camera applications. Some method is used
to get computer control of the focusing and two starting points are
found of some subject. Then a digital picture is taken at incremental
focuses and stored. By looking at areas as either squares, hexagons or
some other regular shape of an area of the picture and calculating the
amount of contrast (by looking at maximum pixel to pixel differences or
other computations method) one can combine the pictures to include the
best contrast for every region of the subject.
The microscope
or camera could automatically gather all the frames of the subject and
the computer combines them in a final picture with perfect contrast. In
some situations it might be necessary to correct for image size changes
and/or field flatness distortions if the focus points are wide.
I've been thinking about the optimizing software - how to use triangles with movable vertexes to set up zones in the image to make the jigsaw out of.
By having movable vertexes the boundaries could be moved to coincide with the objects focus edges so the final image would end up optimized. Would have to do Fourier transform to find the edges. Or better yet take polynomial regressions along a bunch of angled lines - take the derivative of the polynomial to find the rate of change and then identify longest segments. The longest segments of most change would be the edges. Sort of a tomographic approach.
Sounds messy -- scrap the idea in the last paragraph for now - instead first do a transform to compute an image plane that is the difference between the average of the 8 bounding pixels to the center pixel. Then compute an image by grabbing the pixel of the plane with the maximum difference pixels. This would max the contrast at 0,45,90 etc. One could take the pixels marked "o" below at a reduced weight to improve it a bit, but I don't think it would help much.
xxxxxxx xxoxoxx xoxxxox xxx xxx xoxxxox xxoxoxx xxxxxxx
This could be added to current digital cameras as a software feature.
This idea is for a camping ice chest - there would be a box made of Styrofoam that holds the dry-ice. The door to this box would have a bimetallic strip that opens and closes a door on the dry-ice box to maintain a constant temperature in the ice chest without freezing everything and would make the dry ice last much longer.
Dry-ice is getting easier to find these days - the local grocery store has it.
Many people have problems with desiccated or worse disks in their backs that pinch nerves causing back pain. In light of the research showing regeneration of heart muscle with the injection of stem cells it occurs to me that it might be possible to inject stem cells into the disks. Another idea to test would be injecting Human growth hormone directly into the disks - the limited circulation and perhaps some slow release agent might help it stay around long enough to help. Wonder if anyone has injected silicone into disks?
Turns out both of these ideas are being investigated and show they work in rats and rabbits!
Attempts have been made to photograph at very low altitudes from a balloon so as to not make noise or disturb the wild life, but with out good results. A balloon is a very massive aircraft and changing climb descent rates takes time to react so the basket often hits the ground. This idea incorporates a wench in the basket which lowers the basket and allows the photographer to rapidly change his height while the pilot of the balloon just has to keep the balloon at a few hundred feet. There are tow cables going down to provide redundancy and to prevent spinning.
Using microcode of the processor - the instruction set is changed on a regular basis so they can not plant binary code using a buffer overflow. . All binary code would come from a hard drive that would provide a key at start up that would let both the hard-drive and the processor compute a lookup table for the micro code. The Binary executables on the hard drive would be converted to the current instruction set in the processor - the micro code would be volatile.
| Top Page | wiki Index |