Systems and methods for extracting an image of a physical object constructed of for example bricks are presented. The method and system may detect boundaries and edges of a background using an edge detection operator, perform a perspective transformation calculation to compute a corrected virtual grid that is substantially aligned with the physical object's image, locate a color calibration palette in the digital image and extract color value information for pixels of the color calibration palette, and discern bricks as part of the physical object's image, the discernment being based in part on a determination of the brick's color compared to the color palette and the background color, the discerned bricks forming the extracted image. A computer readable medium may include instructions causing a system to extract an image of a physical object constructed of bricks according to the method.