Field Prediction and Potential Monitoring of Cassava Production
in Nakhon Ratchasima
Y. Yamano and K. Saito
Asia Air Survey, Co., Ltd.
13-16, Tamura-cho, Atsugi-shi,
Kanagawa-ken, 243, Japan
Tel: +081-462-22-1002
Fax : +081-462-22-1281
E-mail : KFHO214@niftyserve.or.jp
Abstract
As a new method for the preservation technique of weathered stone relics, the effectiveness of thermal infrared images have been confirmed.
Because the weathering area is distinguished from the other normal area by the difference of the surface temperature change pattern, it is possible to extract the parts deteriorated due to weathering by using thermal images. This method also have characteristics that are indirect, non-destructive and two-dimensional measurements, which are very important for valuable cultural assets.
In this paper, authors introduce two examples which was carried out under the different observation condition. One is the observation for stone relics carved in a cliff. This thermal environment is directly affected by sunbeams. The other is the case of the indoor observation. The target stone exists in the house where the range of the temperature change is smaller than outside.
The subtraction operation between two images that were acquired at different time in a day has been examined in the both case. It has been shown that thermo-anomalous parts were emphasized by these subtracted images and the distribution of weathered parts were consequently enabled to map.
1. Introduction
A new technique using thermal infrared video images has been examined for the test of soil mechanics of weathered stone relics. Characteristics of this technique are to measure by non-touch and non-destructively for the target. Many stone relics are generally made of tuff in Japan. They have been weathered for a long time. For the preservation of weathered relics, the measurement using thermal images has been adopted as a satisfactory technique.
The estimation of weathering condition by the direct looking to the object and by the hitting test of the schmidt hammer has been done until now. These means need to approach and to touch the target. It is a very clear that they are not available to apply to the very high and unclimbable place. As the direct looking depends on personal technique, it includes much subjectivity. Although the hitting test brings the information of the stone strength to us, this method is not able to display the two-dimensional distribution of the weathering area.
On the other hand, the thermal video camera can take the far and dangerous place, because it is able to zoom in. Moreover, thermal images are able to map the two-dimensional distribution for weathering phenomena.
2. Methods
2-1 The surface temperture change in a weathering area.
The surface temperature change is affected by the surface or inner condition of the object. The area that is covered by splitting show larger thermal change than the normal area. The surface temperature rises with the atmospheric temperature increasing. In the normal area, the heat can be transmitted efficiently into the object. the flaking area makes difficult to transmit the heat toward inside. On the other hand, the area including the standing water in cracks shows usually lower temperature than the normal area in the dayp city, therefore the standing water interrupts the heat conduction into the object. These concepts are shown in Figure.1.
Figure 1 The suface temperature change in a day
2-2 Thermal infrared video
Thermal infrared video used in this study has the wave tenth range from 8,000 mm to 13,000 mm. This range decrease the influence of the reflection from the object and it is most efficient among the thermal range (Figure.2).
Figure 2 The wave length of the thermal video
Performance of the thermal video is indicated in Table. 1. It has three thermal range modes. The L-range that is able to measure from-40 c to 160 C was employed in this study. Also it has five scanning time modes. In this study, the 9.6 seconds scanning time mode was used. After the measurment, thermal data were saved into the SCSI hard disk or 3.5 inch floppy disks. One image size is 512 columns by 240 lines and one pixel has 12 bits.
Table.1 Performances of the thermal video
| Items |
Data |
| Measurement range (°C) |
L range |
-40-160 |
| M range |
100-500 |
| H range |
400-2000 |
| Resolving Power(°c) |
0.05(at 30 °c of black body) |
| Angle of visibility(°) |
30(H) x 28(V) |
| Wave length (nm) |
L range |
8000-13000 |
| M range |
8000-13000 |
| H range |
8000-13000 |
| Sensor |
Hgcdte Stirling cooler Type |
| Scanning time (s) |
0.08, 0.15, 0.30, 0.60, 2.40, 9.60 |
| Display pixel numbers |
512(H) x 480(v) |
| Capacity of memory |
512 x 480 x 16 bits x 4 memory |
| Data recording mode |
MS-DOS |
512 x 240 x12 bits
5 images per a disk |
N88
basic |
256 x 230 x 8 bits
15 images per a disk |
2-3 Content of the method
The flow of this method is indicated in Figure. 3. The content of each flow elements is described as follows.
Figure.3 The flow chart of this method
(1) Acquisition of thermal image data
Thermal images were taken two or four times in a day. In the case of two times, thermal images were taken on the early morning and the afternoon. The early morning temperature is the lowest in a day, and it becomes higher in the just before noon or afternoon. If thermo-anomalous areas are included in the target, the characteristic change of the surface temperature will appear between two images.
Four images were taken in the early morning, the before noon. the afternoon and the night. The surface temperature rises from the early morning to the noon and falls from the noon to the night. The difference of two morning thermal images taken in the early morning and in the before noon indicates the thermal condition of the target at the time of temperature rising. The thermal condition at the time of temperature falling are shown by the difference of evening images that are acquired in the afternoon and night.
(2) Geometric correction
The geometric correction was carried out to correct each image.
(3) Subtractive operation of thermal images
In order to get the difference of the surface temperature, one thermal image is subtracted from the image taken at another time. In the case of using two images, the image of higher surface temperature (just before noon or afternoon image) is subtracted by lower (early morning image). The case of using four images get two subtractive images. One is the morning image that the before noon data is subtracted by the early morning data. The other is the evening image that the afternoon data is subtracted by the night data.
(4) Level slice of thermal image
Subtracted images are sliced by the degree of the temperature change and they are classified. In this study, the interval of temperature slice is about 0.25 C at the time of temperature rising and about 0.10 C at the time of temperature falling.
(5) Addition of thermal images
When four images are used, two images operated the level slice, that is, the morning and evening image are added to emphasize the surface temperature change.
(6) Estimation of the weathering area
The thermal condition of the anomalous area is different from the normal area. The reason is shown as follows.
The case of larger temperature change :
- The surface of the target is covered by different materials like splitting.
- There are cavities in the target.
The case of smaller temperature change.
- The cavity of the target includes the standing water.
- The water is welling up from cracks.
(7) Mapping
Thermo-anomalous parts are extracted and the weathering distribution are mapped.
