Hi, it's better to report your data as average particle sizes and also relative abundance diagrams. You can see my paper; I've got three diagram of them.

http://link.springer.com/article/10.1007/s10854-016-4788-7

You have outlined the standard particle size conundrum for irregular materials. See attached. Further every 2D image that you have of a 3D particle is an artifact with TEM.  Consider a collection of identical monodisperse spheres.  Taking thin sections for TEM will produce a distribution of apparent discs where the largest disc diameter measured is the 'true' spherical diameter.  Usually diameters are expressed as equivalents and in this case it could be equivalent circular diameter (the diameter of the circle containing the same area or number of pixels as your irregular particle).

 Thank you so much Mr.Aref and Mr.Alan for your answer.

May be we can measure diameter in several direction of the grain then take the mean of them as a approximation

The diameter of the particle can be measured using Image J software easily. If you don't access this software then it would be better to ask the operater of the TEM to measure the particle while operating the instrument. This would give you exact size whether it is spherical, rod, triangular or any shape.

For more exact calculation follow the link provided by Alan F Rawle Sir.

Dear Dr. Diman .M .Abdullah 

Peace be upon you

 Measuring of the particle diameter or the particle size of HRTEM nano particle can be done using a computational program like Image J software. Otherwise, it can be determined directly at the moment of taking the image by the technician, or it can be done manually, by measuring the diameter of many particles and taking the average value using a suitable magnification.

Let me point out that there are great differences between the crystallite size, grain size and the particle size. The smallest size is the crystallite size, it can successfully be determined by using the X-ray diffraction. While the particle size can be obtained by imaging (photos) using SEM, FESEM, TEM, HRTEM and others.

Thanks and good luck

 thanks you so much Mr,Ahmed for your reply.