Your question is somewhat ambiguous, because the expression 'pixel value' would normally refer to the numerical value of the spectral measurement corresponding to a particular pixel (which would not typically be expressed in m). So, I assume that your query is really about relating the linear size of the area at the Earth surface to the linear size of the elementary detector on the satellite (which generates a 'picture element' or 'pixel' in an image).

For physical and engineering reasons, the ground area that contributes to the measurement acquired by an individual detector is not a clearly defined circular or square region, due, for instance, to adjacency effects and scattering in the atmosphere. Hence, one normally speaks of 'ground sampling distance' or GSD instead: this is the distance between the centers of neighboring areas on the ground sampled by neighboring detector elements, with the understanding that these may not be independent measurements (or, if you prefer, that adjacent areas sampled by adjacent detectors actually overlap each other).

Determining the GSD for a particular sensor does not really require the user to have a 'method' (though the engineers that build the instrument would use techniques such as Ray Tracing and atmospheric radiation transfer models for this purpose): for quasi-circular, geocentric orbits, the relation between the dimension of the detectors and GSD depends essentially on the engineering properties of the sensor, the optical properties of the telescope in front of the detector, and the altitude of the satellite above that surface.

All you need to do is to read the technical documentation provided by the satellite operator, as this relation is essentially a fixed value. Refer to the description of the instrument payload, such as

http://geo.arc.nasa.gov/sge/landsat/l7.html

in the case of Landsat 7. You'll see that the 'ground resolution', which really is a GSD in this case, is given in m for each spectral band. These values will remain constant throughout the useful lifetime of the satellite, or more specifically as long as it remains at the same altitude.

Remember that these values may be different for each satellite, so you will need to explore the appropriate web sites of NASA and USGS to find similar information for the particular satellites of interest to you.

Lastly, if this does not answer your question, please rephrase it to make it clearer.

You can use ImageJ, this is very popular software for image analysis. There is option to measure objects in pixels on the image and convert to meters or other units. But you need some scale of course. 

Are we talking about a geostationary or a polar orbiting satellite? In general search for "geolocation ".

For Landsat product

In that case open your favourite search engine and type "Landsat geolocation" ;-)

hi not able to find my needed one. Kindly requesting you to send me the content to read on it.

Your question is somewhat ambiguous, because the expression 'pixel value' would normally refer to the numerical value of the spectral measurement corresponding to a particular pixel (which would not typically be expressed in m). So, I assume that your query is really about relating the linear size of the area at the Earth surface to the linear size of the elementary detector on the satellite (which generates a 'picture element' or 'pixel' in an image).

For physical and engineering reasons, the ground area that contributes to the measurement acquired by an individual detector is not a clearly defined circular or square region, due, for instance, to adjacency effects and scattering in the atmosphere. Hence, one normally speaks of 'ground sampling distance' or GSD instead: this is the distance between the centers of neighboring areas on the ground sampled by neighboring detector elements, with the understanding that these may not be independent measurements (or, if you prefer, that adjacent areas sampled by adjacent detectors actually overlap each other).

Determining the GSD for a particular sensor does not really require the user to have a 'method' (though the engineers that build the instrument would use techniques such as Ray Tracing and atmospheric radiation transfer models for this purpose): for quasi-circular, geocentric orbits, the relation between the dimension of the detectors and GSD depends essentially on the engineering properties of the sensor, the optical properties of the telescope in front of the detector, and the altitude of the satellite above that surface.

All you need to do is to read the technical documentation provided by the satellite operator, as this relation is essentially a fixed value. Refer to the description of the instrument payload, such as

http://geo.arc.nasa.gov/sge/landsat/l7.html

in the case of Landsat 7. You'll see that the 'ground resolution', which really is a GSD in this case, is given in m for each spectral band. These values will remain constant throughout the useful lifetime of the satellite, or more specifically as long as it remains at the same altitude.

Remember that these values may be different for each satellite, so you will need to explore the appropriate web sites of NASA and USGS to find similar information for the particular satellites of interest to you.

Lastly, if this does not answer your question, please rephrase it to make it clearer.

Jagalingam, Landsat-5 and -7 satellite images usually have a ground resolution of 30 m per pixel for most bands (thermal band 6 has a spatial resolution of 60 m, the panchromatic band 8 has a resolution of 15 m). I take it you would like to know the meter values in terms of topographic/terrain altitude? This information is not included in Landsat images, which rather consist of "grey values" for the reflectance in each of the visible light and infrared bands. However, the Landsat scenes also come with exact corner coordinates and are usually already georeferenced. You can easily drape them over a second terrain dataset, such as ASTER (30 m spatial resolution), SRTM (90 m), or even LiDAR (sub-meter resolution) and then see what's the altitude range/"topography" for each Landsat image pixel. In addition to the commercial software (e.g., ENVI), there is also some free software out there that can handle such data (e.g., Microdem or VTBuilder).

if you want to know the pixel size and its value in meters then just convert the image into UTM projection and the use image info tool in any software like ERDAS and it will show the image value in meters. 

All infomation on image spatial resolution and other features of Landsat images referred for example in USGS web site in FAQ:

http://www.usgs.gov/faq/categories/10164/3839

or

http://www.satimagingcorp.com/satellite-sensors

I do not think that there is a method which can convert reflectance values to meters. The spectral values are unitless and describes how the given detectable object react with the EMR. By the way why you want to change it to meters. Since meter is unit of length measurement, the reflectance values, as far as my knowledge is concerned, will not help to measure distance.

You can use the Beam-Software for different sensors to determine the pixel size in meters.

1 Pixels = 2.645833E-7 Kilometers

Each satellite related to the object that is going to observe, determines special pixel size. Commonly these values are presented as arcsec. Transferring the arcsec to meter is straightforward related to the object which is observed. As an example in the case of the sun it is ~720 km.

Og course dear you just need to multiplied th pixel by it is size it is meant, if you have 10 pixels and the pixel size is 30m * 30 m and you need to know the total size of the 10 pixels so you need to do :

10 *30 *30=9000 square meter

regrads

Hayder Dibs thanks. I am also searching this type of Ans.