Signal to noise enhancement in a study of cell wall structure of Bacillus subtilis by interactive computation

This paper describes a technique for obtaining values of the width and variation in optical density of sectioned bacterial cell walls by interactive computational methods. Background 'noise' prevents accurate determination of cell wall boundaries from the data in a single scan line but the noise may be suppressed by averaging several consecutive scan lines. although application of the technique is explained for sectioned bacterial cell walls, it is equally valid for similar situations where single line scans of electron micrographs are inadequate for precise determination of measurement.