It is believed that the search for alien artifacts on the moon are not needed, because there are none. Thus we find ourselves in a deadlock Circular logic: no finds, hence there is no searching, hence no finds, and so on. Break this vicious circle and is designed to project SAAM. Remote sensing techniques have been used successfully for archaeological sites on the planet. Is it enough to just such a technique for the detection of artificial structures on the Moon and other planets? Hardly, if planetary scientists think only in terms of natural formations. For example, the fortress-temple Coy Krylgan Kala ancient Khorezm (Uzbekistan, IV century. Century BC — I century AD.) Appeared as an impact crater to the excavations in 1956 (Figure 1). Among the many craters of the moon Coy Krylgan-kala would not be noticeable for planetary scientists.
Fig. 2. Modelling the likely form of the ancient settlement on the moon, of the camera HIRES space station "Clementine" (left). Erosion erases traces of the design on the surface (center), but the processing of the SAAM reveal the rectangular anomaly (right).
The set of lunar images taken by the space probe "Clementine" are available in digital form (lit.10). Previous studies of the lunar SETI (lit.11) used a camera image ultraviolet-visible (UVVIS). The resolution of images UVVIS is about 200 meters. According to the criterion of detection that used Sagan, this authorization is not even enough to detect our own civilization on Earth. So even the Great Pyramid would fit entirely in one pixel of the image. Exploration of the moon at this resolution, apparently, did not reveal any convincing evidence of the existence of man-made structures. On the other hand, high-resolution camera (HIRES) "Clementine" produced images of adequate resolution (9-27 m), but they are much more numerous (about 600,000 images) and thus largely unstudied. The next section discusses the algorithms to automatically scan a large number of images from the HIRES-for potential artifacts.
3.1 The preliminary fractal test
Typically, the structure of natural landscapes is self-similar in a large range of sizes. For example, lunar craters between 10E1 to 10E4 m m have almost the same shape. In contrast to the self-similar natural features, the structure of the artificial objects is expressed in a narrow range of sizes. Consequently, the possible artifacts in an image should appear as anomalies in the distribution of spatial detail by size. In search of such anomalies is the essence of the fractal method M.Steyna and Carlotto (lit.12, 13). Unfortunately, their method requires too many calculations for the treatment of all eligible HIRES-images (about 80,000).
Therefore, for the same purpose used an alternative, simpler algorithm. Let M (r) is the probability distribution of distances between local minima of brightness along the horizontal line of the image. Thus, M (r) describes the distribution of the size of the image detail. On large scales, this function can be approximated by a power function, characteristic of fractals:
this area can be considered statistically abnormal with probability 0.95.