將人類肉眼感覺用電腦來模擬，一直是人類努力的目標，由於人眼感覺相當複雜，所以並不容易模擬。在本論文中，我們設計出一套處理技術，希望使電腦能從一堆圖片中判定一自然景觀影像是否具綠草如茵的特性，使電腦能挑選出人眼看起來為綠草如茵的圖片，藉以節省人力及時間。
首先我們要先定義「綠草如茵」，綠草如茵就是綠草一片濃綠，像鋪上褥席一樣，我們就是根據這定義來設計該程式。我們將一張彩色影像的R.G.B模型轉換成H.I.S模型，然後取出H的矩陣，設定好草地部分的H.I.S影像中的H值範圍，讓程式找到符合草地H特性之sample點，接著利用次取樣簡化H矩陣，將這簡化後的矩陣經過完整判定方法來決定草地區域的邊界，然後計算邊界以下之草地面積是否符合我們要的百分比，以此就可以判斷該圖片是否為綠草如茵。我們對於邊界的尋找相當準確，所以辨識率尚稱不錯，辨識速度也不會很慢。
這個處理技術模擬了人眼對一大片均勻草地的感覺，可以提供人們整理拍照後的相片，方便人們找尋旅遊地點或用在教育用途，還有網路上的搜尋圖片功能之改進，避免輸入關鍵字卻搜尋到過多不符合的圖片，或是可以提供攝影機來監控公園、操場、球場等的草皮或綠地，是否已有光禿或草長的不整齊的地方，對於都市環境整理也有幫助。

“Lu-Tsao-Ju-Yin” is a Chinese idiom describing a pleasant scene with uniformly soft and green pastures. We address the issue of identifying whether a given image can be described as “Lu-Tsao-Ju-Yin” in this paper.
We have developed an algorithm to perform the task including transforming the image into an efficient feature space. The complete procedures will be described in the paper. At first we transfer the R.G.B model of a color image to the H.I.S model and take out the H matrix. Then we set the H range of the grassland. This way makes the program to find the sample which is fit in with the character of the grassland. Second we use the sub sampling method to reduce the H matrix. With this reduced matrix, we use the complete judgment to decide the bound of the grassland. Third we estimate the area of the grassland and see if the grassland area rate is answer to the percentage we choose. Finally we can judge an image if it is “Lu-Tsao-Ju-Yin”. The grassland bound is very accurate. So the experiment results on images are good and the the identification speed is not too slow.
This algorithm can learn the feeling of the human eyes on seeing wide and uniform grassland. The technique can be found useful in applications such as sight-seeing and education. For example, we can choose the place which is with grassland without using our eyes. The search engine to images can also be improved.

1） Geun-Won Yoon, Jeong-Ho Park, Kyoung-Ho Choi, “Land-cover Supervised Classification using User-oriented Feature Database”, Volume 4, 20-24 Sept. 2004 Page(s)：2724 – 2726, Geoscience and Remote Sensing Symposium, 2004. IGARSS '04. Proceedings. 2004 IEEE International

2） James M. Stiles, Kamal Sarabandi, Senior Member, “Electromagnetic Scattering from Grassland”, Volume 38, Issue 1, Jan. 2000 Page(s)：349 – 356 , Geoscience and Remote Sensing, IEEE Transactions on

3） Wilbert Long, III and Shobha Srihar, “Land Cover Classification of SSC Image：
Unsupervised and Supervised Classification Using ERDAS Imagine”, Volume 4, 20-24 Sept. 2004 Page(s)：2707 – 2712 , Geoscience and Remote Sensing Symposium, 2004. IGARSS '04. Proceedings. 2004 IEEE International

4） Amit Singhal and Jiebo Luo, “Probabilistic Spatial Context Models for Scene Content Understanding ”, Volume 1, 18-20 June 2003 Page(s)：I-235 - I-241 Computer Vision and Pattern Recognition, 2003. Proceedings. 2003 IEEE Computer Society Conference on

5） Jiebo Luo, Robert T. Gray, “Sunset Scene Classification Using Simulated Image Recognition”, Volume 1, 6-9 July 2003 Page(s)：I - 37-40, Multimedia and Expo, 2003. ICME '03. Proceedings. 2003 International Conference on

6） Le Lu, Kentaro Toyama, Gregory D. Hager, “A Two Level Approach for Scene Recognition”, Volume 1, 20-25 June 2005 Page(s)688 – 695, Computer Vision and Pattern Recognition, 2005. CVPR 2005. IEEE Computer Society Conference on

7） 周 何編, “國語活用辭典” , 第1412頁, 五南圖書出版公司, 1996年6月

8） 繆紹綱編譯, “數位影像處理” , 第316頁, 台灣培生教育出版股份有限公司,2003年8月

9） John C. Russ, “The Image Processing”, Page(s)348–361, CRC Press, 1995