The　rapid　determination　of　the　mechanical　properties　of　geopolymer　concrete　(GPC)　is　important　for　improving　the　efficiency　of　repair　engineering.　In　this　study,　a　rapid　approach　for　the　determination　of　the　mechanical　properties　of　GPC　based　on　image　processing　technology　(IPT)　was　proposed.　First,　a　mesoscopic　structure　model　of　GPC　with　an　aggregate　content　of　40.0　%　was　established　based　on　IPT　and　the　mesoscopic　parameters　of　GPC　were　determined　in　a　numerical　simulation.　Second,　the　accuracy　of　the　mesoscopic　parameters　of　GPC　was　verified　by　comparing　the　numerical　simulation　results　and　experimental　results　of　GPC　with　50.0　%　aggregate　content.　Finally,　a　rapid　approach　combining　IPT　and　a　numerical　simulation　was　applied　to　investigate　the　effects　of　aggregate　contents　on　the　mechanical　properties　of　GPC　and　an　optimal　aggregate　content　was　predicted.　The　results　show　that　the　approach　based　on　IPT　and　numerical　simulation　can　accurately　predict　the　mechanical　properties　of　GPC　with　an　error　of　less　than　10.0　%.　The　failure　mode　predicted　by　the　approach　is　also　consistent　with　the　actual　failure　mode　obtained　by　the　experiment.　The　results　also　show　that　the　optimum　aggregate　content　predicted　by　the　proposed　approach　is　65.0　%,　which　has　the　highest　compressive　strength　and　is　consistent　with　the　experimental　results.　The　proposed　approach　has　potential　application　prospects　in　rapid　repair　engineering,　due　to　its　non-contact,　accurate　and　efficient　characteristics.