Dear all, I am curious about how CCD works to pick up images... more specifically, I wonder CCD camera can output more than just pixel intensity information of an image? For example, I want to know the "variation" property of a block of an image, for example, a 8x8 block... roughly classified into "smooth" block and "non-smooth" block, two classes... I want to know if CCD can judge this very roughly and output some currents to indicate this? Anagolous to human eyes, human eyes can tell which part of the image is "highly non-smooth" at one glance... can CCD do that? Thanks a lot, -Walalla
Can CCD camera output more than just intensity information of image?
Started by ●November 29, 2003
Reply by ●November 29, 20032003-11-29
"walala" <mizhael@yahoo.com> wrote in message news:bq9c7q$5id$1@mozo.cc.purdue.edu...> I want to know if CCD can judge this very roughly and output some currents > to indicate this? Anagolous to human eyes, human eyes can tell which partof> the image is "highly non-smooth" at one glance... can CCD do that?I don't know of any CCD camera that does on board image processing as you describe, but if there is one I'd expect it to be expensive. There's all sorts of IP software that does what you want to a greater or lesser extent, including mine. regards, Stewart DIBBS ======================================== Developer of the FreePiXCL 4.48 PiXCL 5 and geoPiXCL 5 Image Processing Languages for Windows 2000/XP www.vysor.com Gatineau, Quebec, CANADA ========================================
Reply by ●November 29, 20032003-11-29
walala wrote:> Dear all, > > I am curious about how CCD works to pick up images... more specifically, I > wonder CCD camera can output more than just pixel intensity information of > an image? For example, I want to know the "variation" property of a block of > an image, for example, a 8x8 block... roughly classified into "smooth" block > and "non-smooth" block, two classes... > > I want to know if CCD can judge this very roughly and output some currents > to indicate this? Anagolous to human eyes, human eyes can tell which part of > the image is "highly non-smooth" at one glance... can CCD do that? > > Thanks a lot, > > -Walalla"Retina" means "little net". A retina is more than a collection of sensors. It also has connections between sensors that compare outputs and derives information about differences and commonalties among them. Most of the information passed up the optic nerve is of this derivative type. CCDs have no such interconnections. They do no intermediate processing at all. Given the present state of our knowledge, we can't make very good use of the higher-level processing that the eye does. I expect that the depth of our knowledge and the complexity of imaging arrays will advance together, but there's a long way to go. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●November 29, 20032003-11-29
CCD matrix have some connections between photosites => in case of saturation, part of accumlated charge in the well of one photosite can overflow on neigborhood photosite. Unfortunatly it is a parasitic effect leading to 'smear effect'. Completely uncontrolable regards "Jerry Avins" <jya@ieee.org> a �crit dans le message news: bqam4a$gra$1@bob.news.rcn.net...> walala wrote: > > > Dear all, > > > > I am curious about how CCD works to pick up images... more specifically,I> > wonder CCD camera can output more than just pixel intensity informationof> > an image? For example, I want to know the "variation" property of ablock of> > an image, for example, a 8x8 block... roughly classified into "smooth"block> > and "non-smooth" block, two classes... > > > > I want to know if CCD can judge this very roughly and output somecurrents> > to indicate this? Anagolous to human eyes, human eyes can tell whichpart of> > the image is "highly non-smooth" at one glance... can CCD do that? > > > > Thanks a lot, > > > > -Walalla > > "Retina" means "little net". A retina is more than a collection of > sensors. It also has connections between sensors that compare outputs > and derives information about differences and commonalties among them. > Most of the information passed up the optic nerve is of this derivative > type. > > CCDs have no such interconnections. They do no intermediate processing > at all. Given the present state of our knowledge, we can't make very > good use of the higher-level processing that the eye does. I expect that > the depth of our knowledge and the complexity of imaging arrays will > advance together, but there's a long way to go. > > Jerry > -- > Engineering is the art of making what you want from things you can get. > ����������������������������������������������������������������������� >
Reply by ●November 29, 20032003-11-29
samourai wrote:> CCD matrix have some connections between photosites => in case of > saturation, part of > accumlated charge in the well of one photosite can overflow on neigborhood > photosite. > > Unfortunatly it is a parasitic effect leading to 'smear effect'. Completely > uncontrolable > > regards... I think smear is s a bit of a stretch vis-a-vis adjacent-pixel image processing. It's a far cry from what could be achieved by the parallel processing of, say, motion detection, edge detection, and edge orientation directly in the sensor array. We know what to do to achieve those things. What we lack (so far) are practical ways to do them. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●December 1, 20032003-12-01
HI, Today's "smart cameras" most closely fit your description. DSP's, FPGA's and general purpose processors (Intel, Motorola) have been incorporated into a new generation of cameras, CCD and CMOS, to provide quick, local IP at the camera. There are a number of designs where the "smarts" are located in a PC card or separate box as well. The use of smart cameras is becomming increasingly more popular as many applications do not require a descreet computer, have space limitations, or have specific speed and performance requirements. Costs vary, but many are priced at about what you would pay for a good camera, processor, frame grabber, and software anyway. Generally not as flexible as discreet components, they can be very convenient. It would be best to discuss this with an IP professional or Representative who handles a number of different technologies and knows the ins and outs of each. I Know Scott at 1stVision (used to work for him in the past) would be very helpful. Check out his site. Hope this helps, Larry T.
Reply by ●December 1, 20032003-12-01
walala wrote:> > I want to know if CCD can judge this very roughly and output some currents > to indicate this? Anagolous to human eyes, human eyes can tell which part of > the image is "highly non-smooth" at one glance... can CCD do that?The human eye cannot do this, either. However, the human *brain* can. Analogously, the camera just gives you an image. You can do processing on that image to identify areas with strong high-frequency components. -- Mike Smith
Reply by ●December 1, 20032003-12-01
walala wrote:> Dear all, > > I am curious about how CCD works to pick up images... more specifically, I > wonder CCD camera can output more than just pixel intensity information of > an image? For example, I want to know the "variation" property of a block > of an image, for example, a 8x8 block... roughly classified into "smooth" > block and "non-smooth" block, two classes... > > I want to know if CCD can judge this very roughly and output some currents > to indicate this? Anagolous to human eyes, human eyes can tell which part > of the image is "highly non-smooth" at one glance... can CCD do that?You'd be better off just doing some image processing in software than expecting it to come as part of the hardware for a camera. Ben -- I'm not just a number. To many, I'm known as a String...
Reply by ●December 2, 20032003-12-02
Mike Smith wrote:> walala wrote:>> I want to know if CCD can judge this very roughly and output some >> currents >> to indicate this? Anagolous to human eyes, human eyes can tell which >> part of >> the image is "highly non-smooth" at one glance... can CCD do that?> The human eye cannot do this, either. However, the human *brain* can. > Analogously, the camera just gives you an image. You can do processing > on that image to identify areas with strong high-frequency components.As far as I know, human eyes can do the color difference signal calculation, but not other calculations. Other animals might be different. -- glen
