This thesis presents a new study aiming at constraining the gluon contribution {\Delta G} to the 1/2 nucleon spin. The collinear pQCD theoretical framework, on which it is based, deals with asymmetries calculated from cross-sections for single inclusive hadron in the regime of quasi-real photoproduction {Q^2 < 1 GeV^2} at high hadron transverse momentum {p_T > 1 GeV/c). These calculations are done up to Next-to-Leading order with a foreseen inclusion of Next-to-Leading logarithm threshold gluon resummation, only performed for the unpolarised cross-sections yet. This makes the asymmetries sensitive to the gluon polarisation not only through Photon Gluon Fusion {\gamma* g} but also through resolved {\gamma*} processes such as qg or gg.
The measurement of the asymmetries is performed for all the COMPASS data available from 2002 to 2011 with a polarised muon beam at 160-200 GeV scattered off a longitudinally polarised target of deuteron ( {_6LiD} for 2002-2006) or proton ({NH_3} for 2007 and 2011). The asymmetries are presented in bins of pT and of pseudorapidity {\eta_h(p_T \in[1,4]} with {\langlep_T^2\rangle = 3 (GeV/c)^2}, and {\eta_h \in [-0.1, 2.4]}.
Since the resummation calculations are not completed yet for the polarised case, the measurements are only compared with theoretical calculations using different parameterisation sets of polarised Parton Distribution Functions with a large range of different {\Delta G}. This comparison is then used to evaluate the {\Delta G} of these measurements.
Complementary to this analysis, a study of new tracking detectors, the pixelised Micromegas, is performed. After calibration, it shows promising efficiencies and time and spatial resolutions.