I want to determine the temperature of burning coal particles using two colour-pyrometry (analysis of the color channels in the pixles). That is why I need to have very specific information about how the camera saves the images in the formats .raw .dng and .tiff (considering the 8bit adn 16bit versions of tiff) in order to be able to employ the data in matlab.
Using the function imfinfo() in matlab for the 8bit tiff files displayed the following information:
Format: 'tif'-------
FormatVersion: []-------
Width: 512-------
Height: 512-------
BitDepth: 24-------
ColorType: 'truecolor'-------
FormatSignature: [73 73 42 0]-------
ByteOrder: 'little-endian'-------
NewSubFileType: 0-------
BitsPerSample: [8 8 8]-------
Compression: 'Uncompressed'-------
PhotometricInterpretation: 'RGB'-------
StripOffsets: 298-------
SamplesPerPixel: 3-------
GrayResponseUnit: 0.0100-------
MaxSampleValue: [255 255 255]-------
MinSampleValue: [0 0 0]-------
Thresholding: 1-------
Offset: 8-------
To this information I have the following questions:
- Do the MaxSampleValue: [255 255 255] and the MinSampleValue: [0 0 0] represent the black and saturation level? And if so, then what does the offset of 8 represent exactly?
- As there is no information about a linearization table, I assume that there hasn't been any non-linear transformation applied to the sensor data for storage purposes. So is it safe to assume that the data in the tiff file is linear?
- The image I read in is a 512x512x3 uint8 file. And since the documentation shows that it is a truecolor image, I assume the first matrix is the red one, the second is green and the third is blue. Is this correct?
- Has there been any image processing steps applied to the tiff files? I.e. gamma correction or white balancing?