Morphological Research On The Feature Extraction Of Body's Bare Footprints

The plane bare footprints are the tracks information formed by the sole of foot through the force transmission acting on some rigid carrier in the case of body's standing or walking. The morphological feature of its distribution provided with specific characteristics and relative stability is determined by the shape of foot bones. Meanwhile, the bare footprints information which is stemed from person's overall behaviors connects specifically with human being's biological characteristics and presents a certain regularity. This is not only the advantage of the the morphological feature extraction and analysis but also the difficulty. It is rather obscure that the expert's experience describes the mark and location of the track feature points. In addition, the current identification and examination of tracks are artificial or semi-artificial manners. Therefore, we should study the automatic and reliable technology of the shape analysis in order to procure the stable and accurate surveying value of feature. Thus, the footprints can be analyzed and identified.As the preliminary study of the analysis and identification technology of plane bare footprints, this paper provides a model that can extract automatically the survey collections of digital shape feature based on geometric calculation. Also the theory of the standard deviation analysis is used to discuss the recognition ability and robust of the footprints outline feature.During drawing the image outline, we adopt the approach named Threshold Segmentation. Based on the characteristics about the experimental data, we put forward a new method, which is called Dual Threshold Segmentation, combining the Bimodal Method and OTSU. At the same time, in the implementation of OTSU, we use the gray stretch to increase the degree of contrast, then use OTSU to draw threshold to improve the disadvantage, which is caused where the blocks of black regions appear at the segmentations when objectives and the background is not obvious. It is proved in the experiment that this is a segmentation way which can produce an instant and great effect.As to the feature quantity, except for the general geometry characteristics, this paper investigates the information of the two curves of foot, definites the average curvature to describe the curves'entire shape and assesses the angle changes degree of various points along the curves. When analyzing the experimental results and data, we provide a body bare footmark identification model, which is based on the analysis about the collections and the standard deviation comparations of feature quantity. We do some researchs on the relations between the foot feature quantity and the basic information of body, such as weight, gender, height, etc. As the experiment indicates that the feature quantity procured in this paper may associate with some physiological characteristics of the owner and have potential values.