Abstarct: In dentistry science, dental caries are classified in two categories: primary caries and secondary caries. Primary caries are visible by naked eye and does not require special instruments. But since the secondary caries occur below the restored teeth and is not visible by naked eye therefore to diagnose secondary caries dental radiography images are used. In dental radiography images, secondary caries are seen as radiolucent (dark) areas. Decay because of the loss of minerals from tooth cause discoloration of enamel. In this images, besides caries there are other areas seen as radiolucent which are not caries but related to an optical illusion that called Mach band effect. This illusion occurs due to differences in the brightness of the two areas in the images. In restored teeth in the radiography images, differences in the brightness of enamel and restored area, cause to this illusion occurrence. Secondary caries because of similarities with the Mach band of the radiolucent area, the dentist may be wrong and diagnose teeth with mach band as secondary caries and damage to the patient’s teeth. Therefore it is necessary to provide an accurate method to Mach band detection. In this thesis a two-stage method for diagnosis of Mach band in dental radiography images is proposed that the first step involves separating the teeth with secondary caries. The gray level of radiolucent areas in radiography images is reflected in the histogram. So by selecting a threshold on the area under the cumulative frequency diagram of histogram, teeth with secondary caries are diagnosed. In second step the teeth with Mach band are diagnosed. Since Mach band effect occurs due to differences in the brightness of the two areas in the images and depends on the brightness difference, at first by Laplace operator the difference of intensities are calculated and then by selecting a threshold on histogram of Laplace transform of image, teeth with mach band are diagnosed. The proposed method has been carried out on 100 restored teeth images and results indicate that the proposed method has 82% accuracy rate in secondary caries diagnosis and 78% in Mach band detection