Good eye! The second daughtercard connector would more likely be used for a networking card, a 4-20ma output, or some other form of interface appropriate to the application. As it stands, the PropCAM will only work in the righthand socket, since it relies on access to A0-A3 for its data lines, and that socket provides A0-A7. However, with an appropriate adapter, two PropCAMs in one system are definitely possible. The adapter would route A0-A3 to one of them and A4-A7 to the other, with the control signals coming from the lefthand socket. The capture program would then read A0-A7 (two cameras simultaneously).
Simon said...
How are you going to make these available -- thorugh Parallax?
The Parallax team currently has two PropCAM protos to play with and review if they like (and when they have time). Whether they ultimately carry it or not is entirely at their discretion, of course. It's not my decision alone, nor one they will make without due consideration. In addition to any sales potential the PropCAM may have, there's the issue of support costs and how well it complements their current product line. The best new product offerings will not only stand on their own, but amplify the perceived value of existing products.
As a one-man enterprise with minimal overhead, I can operate more in "spaghetti" mode. Cook up a new design; throw it at the wall; if it sticks, great; if not, move on. I don't have to drag an entire development staff along for the ride each time. On the other hand, without employees, there's a limit to how much I can support on my own, without reaching saturation and having no more time to design new stuff. So working with Parallax confers obvious synergistic advantages. Plus the supportiveness they accord their customers also carries over to their suppliers. They're just plain nice to work with.
In any event, I'm moving ahead to make in initial batch of 100 here in the 'States, and have parts available for another 100. If they're well-received, I'll make more and progress towards getting them made overseas.
See the Hydra Status thread http://forums.parallax.com/showthread.php?p=603351 on the discussion in turning a "device" into a "product" for some additional information on the struggles involved for a "one man shop". Good luck Phil, I hope this takes off for you!
Bean (Hitt Consulting) said...
Phil,
Wow, that's alot smaller than it looks in the pictures. Maybe I'll use the Parallax mini servos ? The fixture is going to be huge compared to the camera board.
This is not a bad thing. I need to have multiple sensors on at least one of my sensor turrets, and I want a camera on there also. One of these sensors will be a Devantech Thermal Array Sensor (TPA81) and I think it would be great with a camera, IR, and ultrasonic (SRF08) sensors.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I can handle complexity.. It's the SIMPLE things that confound me.
The Dynaplex Network http://www.thedynaplex.org
The lenses worked out fine, and I'll be ordering a batch soon.
I'm now in the process of panelizing the boards so they can be assembled by a pick-and-place machine, whereupon I'll submit the Gerbers to several fab houses for quotes.
Excellent news! One of the areas of interest for me is image recognition and general vision. It turns out that a black and white camera may be better for image recognition than color. Is this true? If so, I will definitely get in line for at least one PropCam. Actually, I will likely buy at least one regarless. You are doing awesome work and I am so glad you are sharing what you know so freely. I am learning new things from each of your posts and sometimes suffer from information overload.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I can handle complexity.. It's the SIMPLE things that confound me.
The Dynaplex Network - Home of The Octabot Project http://www.thedynaplex.org
Post Edited (linuxgeek) : 9/4/2006 12:06:43 AM GMT
Hi Phil · I know it's been a while, but I have been keeping an eye out on this thread.· So When? I still want, (Need), two PropCams for stereo vision. And I am interested whether parallax is going to decide to carry the Propcam or not! · In reading one or two of the posts a while back. ·I had read something about FRAMs from Ramtron, for use·in place of SRAMs. I did some research and, Well, I bought a few, part#FM18L08s, (32Kx8bits parallel I/O 28pin Dip 3.3v logic). So far, So good! The FRAM Interface uses up most of the Propeller’s I/O, (I’m using the Prop Stick). The FRAM ·address buss is A0-A14 on (Prop pins A17-A31), and the Data Buss is DQ0-DQ7 on (Prop pins A0-A7), the three control lines WE, OE, CE are connected to (Prop pins A8-A10). As an output I’m using the standard (Prop Video pins A12-A15). I’ve started writing the I/O Prop app to go with it. The I/O timing should be fairly simple to implement. I will post some pics ASAP, and I will post the I/O code soon as it's done. (maybe a few day though; Day job family and all). Also, I·bought a few FRAMs 32Kx8bit· with I2C interface only it's at 5volt logic. Can that be interfaced, lets say in place of the EEProm?··
········· So, All of this is leading up to Extra Mem to store images from the PropCam/s. Also, I am starting work on·a new IAI Machine using the Prop. I·call it, "COGnitions". It will be based on the notion of Machine intelligence without programming the program to do the learning. (I.E.. Instead of writing a program that mimics the way the Brain thinks, you only program the way the simplest physical structures work, and·then groups of them form a higher level object, and then groups of them form the next Higher level Object, and so on, I figure (3-7 Cogs deep),·and this level should be Intelligent, (IAI). At lest be able to, "think·like",·infant primates do!) · Like I said, It's going to take some time and a bit more research, but I think I've got the Basics down. So now It's just a matter of·continuing learning spin & asm. and then to finish the translation·from paper into Prop code. · I will keep you informed of the progress. Please let me know when the PropCam/s become Available. · (I welcome all thoughts and suggestions from all Forum members.) · Thanx, IAI Capitan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ IAI (Indigenous Alien Intelligence), The New View on Machine Intelligence. Because There is nothing Artificial about it!
Thanks, really looking forward to doing some vision processing with a propeller.
Seth Vanderdrift
Phil Pilgrim (PhiPi) said...
In any event, I'm moving ahead to make in initial batch of 100 here in the 'States, and have parts available for another 100. If they're well-received, I'll make more and progress towards getting them made overseas.
'Didn't mean to ignore your question for so long. I've been busy helping Parallax get their MoBoStamp-pe ready for sale, so the PropCAM has taken a backseat for awhile. I was also waiting for Parallax's MoBoStamp announcement so it would be easier to convey the big picture encompassing motherboards, daughterboards, BASIC Stamps, Propellers ... and the PropCAM which, in reality, is just a Propeller-specific daughterboard. There is still much to do to flesh all this stuff out, but the PropCAM is integral to my involvement in the overall scheme, and I shall be moving forward with it again very soon.
Phil Pilgrim (PhiPi) said...
Seth,
" ... and the PropCAM which, in reality, is just a Propeller-specific daughterboard. "
-Phil
So thats why you have the ribbon cable on the PropCam! Just rearrange the pins to fit the MoBoProp daughter card socket.
Cool. So I don't have to roll a custom PCB to have an app in the field.
I don't suppose they have a Prop Daughter card coming out. So I could have that hooked to the MoBoProp daughter card socket, and the PropCam hooked up to the daughter card Prop. Have the MoBoProp propeller as the main processor, and the prop daughter card running the PropCam. (with a cog to talk to the MoBoProp processor).
I'm sure someone will come out with a Prop daughter card, but maybe not parallax.
Yes, I have a lens picked out. It's not perfect for all apps (no lens ever could be), but it's a reasonable compromise and cheap enough not to be the tail that wags the dog.
just a random thought.· I understand that the PropCam's speed at capturing data is limited because it can't pack the data into 32-bit words and increment the hub address pointer in two instructions?· How about using one of the timers to generate the address pointer?· This would require·the timer·to increment by 4 every 16 clock-cycles, and the Cog in question would need to be syncronised with the hub to insure that the right address was used.· (or the pointer to the video data structure·could be updated after a transfer to compensate for any jitter in the transfer.··ofc the data structure would need some padding to prevent over-writing something important)
The PropCAM's capture speed is fixed by the sensor hardware at 10MHz; it's not limited by software constraints. Once a capture has begun, it cannot be paused. Each nybble captured requires two instructions: a MOVI to transfer it and a SHR to pack it into cog RAM. So there's no time during the actual capture to write it to the hub. Therefore, two cogs are needed, which alternate between capturing a line of data and formatting and writing the packed LONGs to hub RAM.
Nonetheless, using a timer as an auto-incrementing (or -decrementing) hub pointer is quite an interesting concept, and I'm glad you brought it up. Theoretically, by using two timers and one I/O pin, one could increment at any multiple of twice the clock period. (At least I think that's right.) Initializing it might be a little tricky, but I think it could work.
I'm not entirely sure I understand the question. But the grayscale display object used in conjuction with the PropCAM (though not strictly required by it) is designed to work with a TV driver, not a VGA driver. It needs the analog output in order to produce a grayscale image. That said, it should coexist peacefully with nearly any TV driver, since it merely piggybacks onto an extant signal and doesn't have to generate any syncs on its own. (I haven't actually tried it with PAL, though, to see if it locks on properly.)
Does that clarify things, or did I misss your point completely?
I'm not entirely sure I understand the question. But the grayscale display object used in conjuction with the PropCAM (though not strictly required by it) is designed to work with a TV driver, not a VGA driver. It needs the analog output in order to produce a grayscale image. That said, it should coexist peacefully with nearly any TV driver, since it merely piggybacks onto an extant signal and doesn't have to generate any syncs on its own. (I haven't actually tried it with PAL, though, to see if it locks on properly.)
Does that clarify things, or did I misss your point completely?
-Phil
Phil,
It could actually use VGA if, rather than have two pins each for R, G, and B, you use them all in a single DAC. You could probably get monotonicity at 4 bits using the R+2R+4R+8R topology, but an R-2R would certainly give you six bits (with more resistors).
Somewhere along the way I thought you where using VGA. But, Once again my ignorance has sparked your creative bone. I am glad I asked that though, my demo board was going to sacrifice the TV output. Back to the drawing board!!!
Good luck Phil, We appreciate what you are doing.
And thanks to you also Chip, I have an 8 channel DAC that I will possibly use later, I am designing my board to use the PropDemoBoard version for now so I can better understand the concept.
Everybody's been wondering, "When? When? WHEN???" I can't blame you: it's been a long wait. So here's what's happening:
The lenses arrived today from China, so I can check that off the list. (They were certainly packed in an interesting fashion: they used the plastic lens holders as cushioning — just dumped in the box like plastic peanuts! Amazingly, they all survived unscathed.)
I still need to get some PCBs made, and will likely order them next week. I know, I know! I've really been dithering on this one: whether to have a proto place like Advanced Circuits do a batch first, or go right to production boards from Lazer-Tech in Ontario. If the latter, I've got to invest almost $1000, including NREs, to get a decent price. If the former, the initial cost will be lower, but there'd ultimately be another set of NREs to pay when I switch to production panels.
Another issue is that the new TCS230 color module has had to take precedence over everything else, and I need to get the assembly house up to speed on that before I can even think about getting them building PropCAMS. But I may try to assemble a small batch myself in the meantime, just to get them in people's hands. (There aren't very many parts on it, and I've already made a fixture for soldering the sensor. But if I do it, they won't be lead-free.)
Also, the necessary connecting cables are going to be assembled in China. (DigiKey was laughably expensive.) Parallax is handling that, since they're the same cables used as daughterboard extenders. Then there are the adapter plugs for the Demo Board and PropSTICK. A simple item, but they have to be fabbed and assembled, too (though they could easily be done as kits, since all the parts are through-hole).
Yet, despite it all, I could have still a few units available by year's end. Let's hope!
The camera is not an off-the-shelf item: I'm having it made and have to buy all the parts for it myself. It will come with a lens. As to price: don't know yet. Delivery: soon.
William Chan said...
Why don't you buy the camera like they did with CMU-Cam?
Wouldn't it be easier and cheaper?
The sensor that I'm using has some unique characteristics that make it more suitable for machine vision and which are not found in off-the-shelf video cameras. Most video cameras use what is called a "rolling exposure", in which the integration time for each line begins when the line is output. In the sensor I'm using all the pixels are exposed simultaneously, which yields a truer instantaneous frame acquisition. Also, when you use a standard video camera, you're at the mercy of the camera's output timing and synchronization. By using a sensor that's directly controlled by the Propeller, I can control exactly when each exposure begins and ends. Lastly, the image resolution of the sensor I'm using (128 x 96) is more compatible with the memory available in the Propeller than the higher resolutions of off-the-shelf video cameras.
Phil Pilgrim (PhiPi) said...
Lastly, the image resolution of the sensor I'm using (128 x 96) is more compatible with the memory available in the Propeller than the higher resolutions of off-the-shelf video cameras.
-Phil
Phil,
What's the latest on the PropCam?
Also, what are your thoughts on the future of PropCam with the 'new' proposed 16 cog propellers with 128k ram?
Stereo Vision? Seems like you(we) will have mips to spare.
Comments
Looks like summer's already over here in the UK, so if you're still getting the warm sunny stuff GO TAKE A BREAK!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Cheers,
Simon
Good eye! The second daughtercard connector would more likely be used for a networking card, a 4-20ma output, or some other form of interface appropriate to the application. As it stands, the PropCAM will only work in the righthand socket, since it relies on access to A0-A3 for its data lines, and that socket provides A0-A7. However, with an appropriate adapter, two PropCAMs in one system are definitely possible. The adapter would route A0-A3 to one of them and A4-A7 to the other, with the control signals coming from the lefthand socket. The capture program would then read A0-A7 (two cameras simultaneously).
-Phil
How are you going to make these available -- thorugh Parallax?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Cheers,
Simon
The Parallax team currently has two PropCAM protos to play with and review if they like (and when they have time). Whether they ultimately carry it or not is entirely at their discretion, of course. It's not my decision alone, nor one they will make without due consideration. In addition to any sales potential the PropCAM may have, there's the issue of support costs and how well it complements their current product line. The best new product offerings will not only stand on their own, but amplify the perceived value of existing products.
As a one-man enterprise with minimal overhead, I can operate more in "spaghetti" mode. Cook up a new design; throw it at the wall; if it sticks, great; if not, move on. I don't have to drag an entire development staff along for the ride each time. On the other hand, without employees, there's a limit to how much I can support on my own, without reaching saturation and having no more time to design new stuff. So working with Parallax confers obvious synergistic advantages. Plus the supportiveness they accord their customers also carries over to their suppliers. They're just plain nice to work with.
In any event, I'm moving ahead to make in initial batch of 100 here in the 'States, and have parts available for another 100. If they're well-received, I'll make more and progress towards getting them made overseas.
-Phil
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
John R.
8 + 8 = 10
This is not a bad thing. I need to have multiple sensors on at least one of my sensor turrets, and I want a camera on there also. One of these sensors will be a Devantech Thermal Array Sensor (TPA81) and I think it would be great with a camera, IR, and ultrasonic (SRF08) sensors.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I can handle complexity.. It's the SIMPLE things that confound me.
The Dynaplex Network
http://www.thedynaplex.org
Excellent news! One of the areas of interest for me is image recognition and general vision. It turns out that a black and white camera may be better for image recognition than color. Is this true? If so, I will definitely get in line for at least one PropCam. Actually, I will likely buy at least one regarless.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I can handle complexity.. It's the SIMPLE things that confound me.
The Dynaplex Network - Home of The Octabot Project
http://www.thedynaplex.org
Post Edited (linuxgeek) : 9/4/2006 12:06:43 AM GMT
·
I know it's been a while, but I have been keeping an eye out on this thread.·
So When?
I still want, (Need), two PropCams for stereo vision.
And I am interested whether parallax is going to decide to carry the Propcam or not!
·
In reading one or two of the posts a while back. ·I had read something about FRAMs from Ramtron, for use·in place of SRAMs. I did some research and, Well, I bought a few, part#FM18L08s, (32Kx8bits parallel I/O 28pin Dip 3.3v logic). So far, So good!
The FRAM Interface uses up most of the Propeller’s I/O, (I’m using the Prop Stick). The FRAM ·address buss is A0-A14 on (Prop pins A17-A31), and the Data Buss is DQ0-DQ7 on (Prop pins A0-A7), the three control lines WE, OE, CE are connected to (Prop pins A8-A10). As an output I’m using the standard (Prop Video pins A12-A15).
I’ve started writing the I/O Prop app to go with it. The I/O timing should be fairly simple to implement. I will post some pics ASAP, and I will post the I/O code soon as it's done. (maybe a few day though; Day job family and all).
Also, I·bought a few FRAMs 32Kx8bit· with I2C interface only it's at 5volt logic. Can that be interfaced, lets say in place of the EEProm?
········· So, All of this is leading up to Extra Mem to store images from the PropCam/s.
Also, I am starting work on·a new IAI Machine using the Prop. I·call it, "COGnitions". It will be based on the notion of Machine intelligence without programming the program to do the learning. (I.E.. Instead of writing a program that mimics the way the Brain thinks, you only program the way the simplest physical structures work, and·then groups of them form a higher level object, and then groups of them form the next Higher level Object, and so on, I figure (3-7 Cogs deep),·and this level should be Intelligent, (IAI). At lest be able to, "think·like",·infant primates do!)
·
Like I said, It's going to take some time and a bit more research, but I think I've got the Basics down. So now It's just a matter of·continuing learning spin & asm. and then to finish the translation·from paper into Prop code.
·
I will keep you informed of the progress.
Please let me know when the PropCam/s become Available.
·
(I welcome all thoughts and suggestions from all Forum members.)
·
Thanx,
IAI Capitan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
IAI (Indigenous Alien Intelligence),
The New View on Machine Intelligence.
Because There is nothing Artificial about it!
Any updates on the status of the PropCAM?
Thanks, really looking forward to doing some vision processing with a propeller.
Seth Vanderdrift
'Didn't mean to ignore your question for so long. I've been busy helping Parallax get their MoBoStamp-pe ready for sale, so the PropCAM has taken a backseat for awhile. I was also waiting for Parallax's MoBoStamp announcement so it would be easier to convey the big picture encompassing motherboards, daughterboards, BASIC Stamps, Propellers ... and the PropCAM which, in reality, is just a Propeller-specific daughterboard. There is still much to do to flesh all this stuff out, but the PropCAM is integral to my involvement in the overall scheme, and I shall be moving forward with it again very soon.
Thanks for your patience and continued interest!
-Phil
So thats why you have the ribbon cable on the PropCam! Just rearrange the pins to fit the MoBoProp daughter card socket.
Cool. So I don't have to roll a custom PCB to have an app in the field.
I don't suppose they have a Prop Daughter card coming out. So I could have that hooked to the MoBoProp daughter card socket, and the PropCam hooked up to the daughter card Prop. Have the MoBoProp propeller as the main processor, and the prop daughter card running the PropCam. (with a cog to talk to the MoBoProp processor).
I'm sure someone will come out with a Prop daughter card, but maybe not parallax.
Did you get the lens problem worked out?
Seth
Hmm. 'Never thought about a Prop daughtercard...
Yes, I have a lens picked out. It's not perfect for all apps (no lens ever could be), but it's a reasonable compromise and cheap enough not to be the tail that wags the dog.
-Phil
Yea, done 'properly', the daughter card could work with the MoBoStamp or MoBoProp. Althought using it with the MoBoStamp, just 'feels' wrong.
Seth
Sincerly,
Marty
The PropCAM's capture speed is fixed by the sensor hardware at 10MHz; it's not limited by software constraints. Once a capture has begun, it cannot be paused. Each nybble captured requires two instructions: a MOVI to transfer it and a SHR to pack it into cog RAM. So there's no time during the actual capture to write it to the hub. Therefore, two cogs are needed, which alternate between capturing a line of data and formatting and writing the packed LONGs to hub RAM.
Nonetheless, using a timer as an auto-incrementing (or -decrementing) hub pointer is quite an interesting concept, and I'm glad you brought it up. Theoretically, by using two timers and one I/O pin, one could increment at any multiple of twice the clock period. (At least I think that's right.) Initializing it might be a little tricky, but I think it could work.
Cheers!
Phil
Just checken on ye! Hoping all is well?
Is it close yet ?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Cheers,
Simon
www.norfolkhelicopterclub.co.uk
You'll always have as many take-offs as landings, the trick is to be sure you can take-off again ;-)
Closer. I've ordered 300 lenses and lens holders from China. Hopefully, they have shipped, although I've not received confirmation yet.
-Phil
Does the new video driver Chip came up with give you any regrets about your PropCam Object.
I'm not entirely sure I understand the question. But the grayscale display object used in conjuction with the PropCAM (though not strictly required by it) is designed to work with a TV driver, not a VGA driver. It needs the analog output in order to produce a grayscale image. That said, it should coexist peacefully with nearly any TV driver, since it merely piggybacks onto an extant signal and doesn't have to generate any syncs on its own. (I haven't actually tried it with PAL, though, to see if it locks on properly.)
Does that clarify things, or did I misss your point completely?
-Phil
It could actually use VGA if, rather than have two pins each for R, G, and B, you use them all in a single DAC. You could probably get monotonicity at 4 bits using the R+2R+4R+8R topology, but an R-2R would certainly give you six bits (with more resistors).
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chip Gracey
Parallax, Inc.
Another option, I suppose, would be to use an additional single output in DUTY mode and sum it, through appropriate resistors, into the RGB pins.
-Phil
Somewhere along the way I thought you where using VGA. But, Once again my ignorance has sparked your creative bone. I am glad I asked that though, my demo board was going to sacrifice the TV output. Back to the drawing board!!!
Good luck Phil, We appreciate what you are doing.
And thanks to you also Chip, I have an 8 channel DAC that I will possibly use later, I am designing my board to use the PropDemoBoard version for now so I can better understand the concept.
Everybody's been wondering, "When? When? WHEN???" I can't blame you: it's been a long wait. So here's what's happening:
The lenses arrived today from China, so I can check that off the list. (They were certainly packed in an interesting fashion: they used the plastic lens holders as cushioning — just dumped in the box like plastic peanuts! Amazingly, they all survived unscathed.)
I still need to get some PCBs made, and will likely order them next week. I know, I know! I've really been dithering on this one: whether to have a proto place like Advanced Circuits do a batch first, or go right to production boards from Lazer-Tech in Ontario. If the latter, I've got to invest almost $1000, including NREs, to get a decent price. If the former, the initial cost will be lower, but there'd ultimately be another set of NREs to pay when I switch to production panels.
Another issue is that the new TCS230 color module has had to take precedence over everything else, and I need to get the assembly house up to speed on that before I can even think about getting them building PropCAMS. But I may try to assemble a small batch myself in the meantime, just to get them in people's hands. (There aren't very many parts on it, and I've already made a fixture for soldering the sensor. But if I do it, they won't be lead-free.)
Also, the necessary connecting cables are going to be assembled in China. (DigiKey was laughably expensive.) Parallax is handling that, since they're the same cables used as daughterboard extenders. Then there are the adapter plugs for the Demo Board and PropSTICK. A simple item, but they have to be fabbed and assembled, too (though they could easily be done as kits, since all the parts are through-hole).
Yet, despite it all, I could have still a few units available by year's end. Let's hope!
Cheers!
-Phil
How much $$ to order a unit and When?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.fd.com.my
www.mercedes.com.my
The camera is not an off-the-shelf item: I'm having it made and have to buy all the parts for it myself. It will come with a lens. As to price: don't know yet. Delivery: soon.
-Phil
Thanks for the update!
Hope you can keep me on the short list for any early releases.
Cheers!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
The sensor that I'm using has some unique characteristics that make it more suitable for machine vision and which are not found in off-the-shelf video cameras. Most video cameras use what is called a "rolling exposure", in which the integration time for each line begins when the line is output. In the sensor I'm using all the pixels are exposed simultaneously, which yields a truer instantaneous frame acquisition. Also, when you use a standard video camera, you're at the mercy of the camera's output timing and synchronization. By using a sensor that's directly controlled by the Propeller, I can control exactly when each exposure begins and ends. Lastly, the image resolution of the sensor I'm using (128 x 96) is more compatible with the memory available in the Propeller than the higher resolutions of off-the-shelf video cameras.
-Phil
Phil,
What's the latest on the PropCam?
Also, what are your thoughts on the future of PropCam with the 'new' proposed 16 cog propellers with 128k ram?
Stereo Vision? Seems like you(we) will have mips to spare.
Thanks
Seth
armondo522@hotmail.com