Granary Stored Grain Volume Measurement In The Laser Point Cloud Data Processing Technology

Warehouse inventory checking of national grain obtains the total grain, finds out the property inventory, and services the structure of grain production, the grain macro-control and the protection of national grain security. It can promote the development of economy steadily and rapidly. In the work of warehouse inventory checking, it's very important to measure the volume of grain accurately and quickly. But it has been a technical problem over the years. The traditional methods always calculate the volume by using mathematical formulas based on the shape of accumulative grain in the granary, which not just wastes a lot of human and material resources, but can't acquire the number fast. In this paper, under the financial aid of National Science and Technology Support:study of non-contact and fast detection technique of granary reserve (2008BADA8B02-1), the methods of calculating the volume is developed, applying 3D laser scan technology on detection of granary reserve. First, the author obtains the point cloud data of granary and grain surface using 3D laser scanner, then acquires the complete point cloud data of grain after data registration and integration. According to the shape characteristics of point cloud data, the author designs the projection and slicing methods which can obtain the volume of grain in the granary directly and efficiently. The algorithm was successfully implemented in the software, whose feasibility and correctness were verified by calculating the experimental data. The author makes some innovation achievements on the following aspects:1) Data reduction:In order to accelerate the speed of the algorithm and improve the accuracy of point cloud data, the measured mass-data need to be de-noise and reduction;2) Data registration:According to the geometric information of granaries, the common features of the data sets can be extracted. If the features can't be extracted accurately, the author extracts the feature points interactively. Then, the data registration will be completed by using ICP algorithm. Finally, the strategy of global registration was developed:building a base station in the center of granary, the others align to the base site collectively, which avoids the cumulative error.3) Data integration:There will be a lot of redundant data in the overlap region after registration. So the article takes the mean distance sampling methods to remove the redundant data. If the point cloud data is not complete, the linear interpolation or quadratic surface interpolation will be used to repair the data.4) Volume calculations:According to the shape characteristics of granary, the article will choose the projection method to calculate the volume of data if the perspective plane can be determined easily. If the data is closed, the slicing method will be selected.5) Experimental Verification:The article calculates the volume of measured granary. In order to verify the accuracy of the results, the accuracy of 3D laser scanner and the algorithm of volume calculation are verified. At last, the article verifies the total accuracy.