An image processing apparatus includes a memory configured to store a data structure in which gather rays generated by performing gather ray tracing from a rendering viewpoint are clustered hierarchically, the gather rays including a first gather ray and a second gather ray associated with the first gather ray in the data structure, and a processor configured to merge a photon gathering result associated with the second gather ray with a photon gathering result associated with the first gather ray in response to the photon gathering result associated with the first gather ray not satisfying a first condition.