ANN might be used to improve psychiatric diagnosis.

No abstract

No abstract

As large scale complex surface product encounters problems such as large point cloud data density, various redundancy points and heavy processing tasks in reverse engineering, the Thesis analyzes necessity of data pre-processing in reverse engineering. For dispersing of data measurement, coarse point cloud becomes smooth and level through point cloud pretreatment and there are only a few miscellaneous points and the gained point cloud reduces operating difficulty of three-dimensional model reconstitution significantly; the Thesis proposes final judgment criterion () S S ′ ∆ , during point cloud sampling. The judgment criterion meets requirement of high precision, fast speed and suitable simple principles and it can also reach the given maximum permissible error or quantity of the specified points and it applies to occasions having high requirement on geometric accuracy of reconstruction; at last, the Thesis takes certain large scale hydraulic turbine blade as an example and conducts sampling and reverse design modeling accuracy analysis and proves validity and feasibility of three-dimensional optical detection point cloud sampling method. Point cloud pretreatment technology Point cloud pretreatment procedure According to Fig.1, point cloud pretreatment technology procedure includes acquisition of large quantity of point cloud data, point cloud pretreatment and surface reconstruction. Pretreatment technology is the core and difficulty of the research and it includes: