Brown planthopper (BPH) is one of the most dangerous kind of insect for rice plant. Detecting and estimating BPH population in rice field are concerned recently to support the BPH monitoring of this kind of insect. In this paper, we introduce an approach to detecting and counting BPH in images taken by light traps, which is the most used device to catch BPH in rice fields. Our approach uses morphological operations to analyse BPH structure and the sequential region labelling algorithm to count the separate BPHs. Moreover, it is common that the BPH in light trap images are overlapped due to the layout of light traps. Therefore, an approach to dealing with this problem is also proposed. The experimental results show that our proposed approach is suited to identify and count the BPHs in images even in case that they contain the overlapping individuals.