Study On Retrieving Of Point Map Symbol Based On Shape Features

The map is a very important kind of information resources, extremely wide range of applications in many fields of economic construction, scientific theories and research. The composition of the map symbol is part of the point, line, surface, and therefore, for the identification of the map, in fact, recognizing symbols of these three elements. The identification of point symbols is the foundation of map identification.During the process of identifying objects, the physical characteristics and structural features are used more often, since people like to understand the features of an object with the eyes to see and ears to hear, or other sensory means. For a computer, now we can not simulate the feeling, so it is unlikely to make the computer to identify these specific sensory characteristics. The computer’s expertise is extracting the mathematical characteristics which is the human senses can not do. Therefore, the use of the mathematical characteristics instead of describing objects in the sensory characteristics of an important basis for the recognition of object features.The moment invariants is a kind of overall situation image features, it has the RST invariance which can shape characteristics of the image in the region, so we can use it to identify the characteristics of the point map symbol. The main content of the paper include:1. Comprehensive description of domestic and foreign point map symbol recognition, and introduces the technical means of content-based image retrieval theory, and analyzes the major technical problems exist in the content-based image query.2. Analysis of the main features of the point map symbol, the composition discipline, based on moment invariant features of point map symbol feature extraction with RST invariance.3. Detailed the study of the theory of invariant moments to analyze the moments of actual physical significance of RST invariance of the derivation and validation, and the two invariant moments-HU moments and Zernike moments were introduced including their invariance derivation and proof.4. Various moment features, the use of professional and scientific computing software Matlab extraction point map symbols, map symbols, similarity matching. The experimental results demonstrate that the use of moment feature point map symbol matching method is feasible, and its precision and accuracy to the actual application requirements.5. The invariant moment matching results, and retrieval operations on the actual map symbols. Experiments show that the use of invariant moments for the map symbol to retrieve the effect can reach a certain level.