On the research of evaluating optical camouflage ground targets, a method using tolerance nearness sets to evaluate ground targets′ camouflage effect is proposed, according to the characteristics of the camouflage images. The targets and backgrounds are extracted from these images, which seen as different sets in tolerance space, and divided into numbers of sub-images. Several features such as statistics, color and texture characteristics are extracted from every sub-image and all of tolerance nearness class in background and object sets are calculated. Nearness measure (tNM) is used to evaluate camouflage effect and compared with Hausdorff distance (tHD). The results show that tolerance nearness sets method can well replace human eyes to evaluate camouflage effects and tNM is better than tHD as the evaluation index.