評估數位影像技術量測大氣能見度指標之實驗方法

An experiment method of evaluating the digital image processing schemes for measuring atmospheric visibility index

蔡岱玲(Tai-Ling Tsai)

指導教授 ： 廖俊鑑

朝陽科技大學/資訊學院/資訊與通訊系/碩士(2014年)

摘要

大氣中的能見度(Atmospheric visibility)跟我們的生活有密切的關係，能見度高低就表示空氣品質佳與不佳，進而影響到交通安全及人體健康。以往的能見度量測大多以人為觀測，雖然觀測者受過專業訓練，但還是有可能因為個人主觀問題，而使測量的結果產生誤差。
過去研究有提出使用數位影像處理來測量能見度，所以本論文提出一個新的實驗環境。利用理想高通濾波器(ideal high-pass filter)、同態濾波器(homomorphic filter)、哈爾函數(haar function)和灰階對比法(itccb)。利用上面四種方法來進行能見度的測量，在實驗中我們把霧氣視為空氣中會影響能見度的懸浮微粒，利用霧氣使玻璃箱的光照度改變，模擬霧氣改變產生能見度的變化。實驗將光照度和能見度指標(index value)作線性回歸(linear regression)，利用兩者的相關性計算出相關係數(R2)，藉由相關係數統計圖證明使用機器量測的R2值會符合之前學者的實驗結果。

關鍵字

能見度指標 ； 大氣能見度 ； 線性回歸

並列摘要

Atmospheric visibility is closely related with our live. Low visibility means the air quality is not good. It will consequently affect the traffic safety and the human health. In the past, the measurement for atmospheric visibility is mostly done by the human observers. Although those observers have been well-trained professionally, there is still a problem due to the observers, personal subjectivity. It causes the errors in the results of measurement.
Previous researches have proposed to use the digital image processing to measure the visibility. Therefore, we propose to create a new experimental environment in this paper. We use the four methods the ideal high-pass filter, the homomorphic filter, the haar function and the grayscale contrast method (itccb) to measure the visibility. In the experiment, we put the mist in the air as suspended particles to affect the visibility, use the mist in glass box to change illumination, stimulating the increase of the mist to record the changes of the visibility. We do the experiment of the illumination and the visibility index for linear regression. We use the correlation between the illumination and the visibility index to calculate the correlation coefficient (R2). The correlation coefficient charts proves the value will be conform by using the machine-measured (R2) than that before by the scholars.

並列關鍵字

index value ； Atmospheric visibility ； linear regression

參考文獻

[4]. R. Sequeira and K. H. Lai, “The effect of meteorological parameters and aerosol constituents on visibility in urban Hong Kong,” Atmospheric Environment, pp. 2865-2877, 1998.

[5]. H. Horvath, “Atmospheric visibility,” Atmospheric Environment,pp. 1785-1796. , 1981

[6]. S. M. Larson, G. R. Cass, K. J. Hussey, and F. Luce, “Verification of image processing based visibility models,” Environmental Science and Technology, pp. 629-637, 1988.

[7]. C. H. Luo, S. H. Liu, and C. S. Yuan, “Measuring atmospheric visibility by the digital image processing,” Aerosol and Air Quality Research, pp. 23-29, 2002.

[8]. M. L. Pitchford and W. C. Malm, “Development and application of a standard visual index,” Atmospheric Environment, pp. 1049-1054, 1994.