The background correction is designed to enhance the detectability of a substance that has a narrow absorption band in the visible region compared with those typical of most substrates. The method is illustrated in this paper by application to the detection of blood by using a two-dimensional extension of the Allen correction of blood quantitation. The method described is designed to visually identify regions where blood is likely to be found on a surface whose texture, color, and/or lighting conditions make it difficult to spot areas that warrant testing for blood. This paper describes the procedure for determining whether a particular digital camera might be suitable for generating images for background detection. The limits due to the camera are both related to noise in an individual image and variability between images. Ways of dealing with these limits are discussed. Further, the three-wavelength method of background correction described requires that the pixels in the original photograph are at appropriate brightness levels. The upper and lower limits of pixel brightness must be determined for a particular camera. The authors describe how to make this determination. Should a set of images of a potentially bloodstained item include pixels that are underexposed or overexposed, the authors explain two methods of addressing this problem. The paper describes the camera used in the experiments profiled in this paper, as well as the light source and the processing of the images. 7 figures and 11 references