A method of forming of an image sensor device includes a substrate having a pixel region and a periphery region. A plurality of first trenches is etched in the periphery region. Each of the first trenches has a depth D1. A mask layer is formed over the substrate. The mask layer has a plurality of openings in the pixel region. A spacer is formed in an interior surface of each opening. A plurality of second trenches is etched through each opening having the spacer in the pixel region. Each of the second trenches has a depth D2. The depth D1 is larger than the depth D2.