基于无人机遥感影像的建筑物轮廓目标检测系统设计

Design of building contour target detection system based on UAV remote sensing image

摘要:

使用?CCD成像元件系统和基于动态扫描检测系统受到噪声影响，导致轮廓检测不完整，为解决这一问题，提出了基于无人机遥感影像的建筑物轮廓目标检测系统设计。依据系统总体架构可确定系统分为数据采集模块和处理模块。选择TLC2543A/D转换芯片，使控制字从数据输入终端连续输入。使用半导体激光传感器测量位置信息，在雪崩光电二极管上成像。采用MCS-51单片机内部数据存储器，将正常运行程序传送到高层，以便?CPU读取程序。选择371-4615型号CPU板，协调控制整机。设计轮廓目标检测流程，采用激光三角法测量方法计算待测建筑物轮廓目标距离，避免检测背景噪声影响。提取建筑物斑块，拐角点定位，完成建筑物轮廓目标检测。由实验结果可知，该系统不会受到噪声影响，能够确定建筑物轮廓和拐角，对建筑物轮廓目标识别具有一定参考意义。

Abstract:

The use of CCD imaging element system and dynamic scanning-based detection system is affected by noise, resulting in incomplete contour detection. To solve this problem, the design of building contour target detection system based on UAV remote sensing image is proposed. According to the overall architecture of the system, it can be determined that the system is divided into a data acquisition module and a processing module. Select TLC2543A / D conversion chip to make the control word input continuously from the data input terminal. Use a semiconductor laser sensor to measure position information and image on an avalanche photodiode. Adopt MCS-51 one-chip computer internal data memory, transfer the normal operation procedure to the high-level, so that CPU reads the procedure. Select the 371-4615 model CPU board to coordinate and control the whole machine. Design the contour target detection process, use laser triangulation measurement method to calculate the distance of the contour target of the building to be measured, to avoid the influence of detection background noise. Extract building patches and locate corner points to complete building contour target detection. It can be seen from the experimental results that the system will not be affected by noise, can determine the contours and corners of buildings, and has certain reference significance for the target recognition of building contours.