Perceiving a color difference is different from identifying the colors causing the difference. The former is a subtask of the latter that involves the tasks of color identification and color difference perception. In this paper, a technique capable of detecting specific color edges in an image is proposed. Instead of developing an efficient algorithm for detecting edges caused by specific colors, we prefer a process that can be effectively implemented in hardware. The proposed process first calculates the differences of the individual color components of each pixel. Next, the colors causing the differences are identified by analyzing the calculated differences of color components. An architecture realizing the process using an FPGA-based SOPC is detailed. The architecture is readily parallelizable and allows users to specify the colors that form the edges to be located.