利用近红外彩色成像技术监测松萎蔫病

Detection of pine wilt-damaged trees using near-infrared color photograph

有效监测患病株方法的缺失是导致对松萎蔫病难以成功控制的主要原因之一。利用航空手段调查该病导致的针叶颜色变化已被证明是一种可靠的方法。为了克服该方法固有的缺点,笔者引用了近红外彩色胶片航拍方法。该项技术在获得正摄航拍图像时,通过图像处理可以增强松针颜色变化,获得较直接拍摄图像更明显的色差。因为图像可以在室内仔细处理,因此能解决飞行时间缺乏等问题。但是,该方法也存在对被压木无法获得其影像而漏检的缺陷。笔者研发的计算机软件可以将目标树标记在计算机图像文件上,并且标记树的地理位置数据和背景航拍图像并传输到装有内置GPS接收器的掌上电脑中,借助软件产生的图像导航定位系统,有利于地面接近标记树,从而可以现场检查和校正标记树的数据,并将修正数据传回至主机。此方法利用最新航拍图像技术建立对每株树处理和管理的连续资料,大大改善了防控松萎蔫病的措施。

The failure to detect damaged trees harboring pine wilt disease insect vectors is one of the main reasons for lack of success in the control of pine wilt disease（PWD）.Aerial surveys to detect wilt-affected trees,i.e.those with discolored foliage,has been proven to be a reliable disease survey technique and as such many governments in Japan have already adopted such methods.This conventional method of aerial survey and following the treatment to detect trees,however,suffer from some technical problems such as insufficient flight time for a detailed survey,the ambiguity of foliage coloration,and the imprecise location of the spotted trees causes difficulties in locating them from the ground.To overcome these problems we introduced using near-infrared color films. Recent progress in such technology enables us to readily obtain an ortho-rectified aerial photograph that can provide the exact location （latitude/altitude） of any point.Properly processed near-infrared photos enhance foliage discoloration and make it more distinct than that in natural color photographs. With the photographs, we can carefully search for the discolored trees in the office without flight-time restrictions.Recognition of trees with foliage discoloration is highly accurate and can be done by personnel with minimal training, however,detection is not 100 % precise because of some problems such as the invisibility of shaded trees from above.The photographs may be electronically reproduced and displayed on a computer monitor. We have developed such a software with which we can designate the target trees and obtain their coordinates on an image file shown on the monitor. The location data for the target trees with the background image of the aerial photograph can be transferred to a palmtop computer with a built-in GPS receiver and the other original software will produce a graphical navigation system for ground approach of the target trees. Using the palmtop computer we can check and edit the data of the target trees on site, and the updated data can be transferred back to the host computer.A consecutive system from detecting the trees to be treated to easy access for disposal and management of the individual tree data will be provided by making good use of the latest technology on aerial photography. This will bring an extensive improvement in control operations against PWD.