Figure 2. Technical Route
III. DATA PROCESSING AND ANALYSIS RESULTS
A. Multiple Classified methods of Image Interpretation
It should have the same or similar characteristics of spectral and spatial information when the same features in the similar conditions (light, terrain, etc.). on the contrary, different types of surface features has different diversities. According to this difference, all the pixels of the image are divided into several categories based on their nature properties (Class). Regardless of the features’ informational extraction, the dynamic monitoring or the productions of thematic maps and remote sensing images to establish, they all can do without classification. The specific application of computer classification is to identify and classify the features atribute of remote sensing image and It is also the pattern of recognized technology in the field of remote sensing. The purpose of classification is to identify the actual surface features from the image and then to extract information about surface features, the process is that each pixel point or area is attributed to a number of categories in one class actually or several thematic elements in the one, it will complete the twodimensional gray-scale image data from space to the target pattern space.
According to the classification need, personnel provide the training samples and the known category, training and supervising the classifier, remote sensing image classified method can be divided into supervised classification and nonsupervised classification.
(1) Non-Supervised Classification
The classification process does not impose any priori knowledge, It just depends on the pixel’s spectral characteristics, according to the similarity into several categories, the research uses the Iterative Self-Organizing Data Analysis (ISODATA) algorithm to classify a group of
pixels.
(2) Supervised Classification
Different from unsupervised classification, the supervised classification is on the basis of a priori knowledge to distinguish from various types of surface features’ samples on remote sensing images, then set up templates and take automatic identification. The basic process of image classification as shown below:
Figure 2. Technical Route
III. DATA PROCESSING AND ANALYSIS RESULTS
A. Multiple Classified methods of Image Interpretation
It should have the same or similar characteristics of spectral and spatial information when the same features in the similar conditions (light, terrain, etc.). on the contrary, different types of surface features has different diversities. According to this difference, all the pixels of the image are divided into several categories based on their nature properties (Class). Regardless of the features’ informational extraction, the dynamic monitoring or the productions of thematic maps and remote sensing images to establish, they all can do without classification. The specific application of computer classification is to identify and classify the features atribute of remote sensing image and It is also the pattern of recognized technology in the field of remote sensing. The purpose of classification is to identify the actual surface features from the image and then to extract information about surface features, the process is that each pixel point or area is attributed to a number of categories in one class actually or several thematic elements in the one, it will complete the twodimensional gray-scale image data from space to the target pattern space.
According to the classification need, personnel provide the training samples and the known category, training and supervising the classifier, remote sensing image classified method can be divided into supervised classification and nonsupervised classification.
(1) Non-Supervised Classification
The classification process does not impose any priori knowledge, It just depends on the pixel’s spectral characteristics, according to the similarity into several categories, the research uses the Iterative Self-Organizing Data Analysis (ISODATA) algorithm to classify a group of
pixels.
(2) Supervised Classification
Different from unsupervised classification, the supervised classification is on the basis of a priori knowledge to distinguish from various types of surface features’ samples on remote sensing images, then set up templates and take automatic identification. The basic process of image classification as shown below:
การแปล กรุณารอสักครู่..
รูปที่ 2 ผล
เทคนิคการประมวลผลข้อมูลและวิเคราะห์เส้นทาง III A หลายวิธีของการตีความตามภาพ
มันควรมีลักษณะที่เหมือนกันหรือคล้ายกันของสเปกตรัมและข้อมูลเกี่ยวกับอวกาศเมื่อคุณสมบัติเดียวกัน ในเงื่อนไขที่คล้ายกัน ( แสง , ภูมิประเทศ , ฯลฯ ) ในทางตรงกันข้าม ประเภทที่แตกต่างกันของพื้นผิวมีความหลากหลายแตกต่างกัน ตามความแตกต่างนี้ all the pixels of the image are divided into several categories based on their nature properties (Class). Regardless of the features’ informational extraction, the dynamic monitoring or the productions of thematic maps and remote sensing images to establish, they all can do without classification. The specific application of computer classification is to identify and classify the features atribute of remote sensing image and It is also the pattern of recognized technology in the field of remote sensing. The purpose of classification is to identify the actual surface features from the image and then to extract information about surface features, the process is that each pixel point or area is attributed to a number of categories in one class actually or several thematic elements in the one, it will complete the twodimensional gray-scale image data from space to the target pattern space.
According to the classification need, personnel provide the training samples and the known category, training and supervising the classifier,ภาพลับวิธีการระยะไกลสามารถแบ่งออกเป็นหมวดหมู่และประเภทการ nonsupervised .
( 1 ) ไม่มีกระบวนการจัดหมวดหมู่
ไม่กําหนดใด ๆระหว่างความรู้ มันก็ขึ้นอยู่กับลักษณะสเปกตรัมของพิกเซลตามความคล้ายคลึงกันในหลายประเภทการวิจัยใช้วิธีการด้วยตนเองการจัดวิเคราะห์ข้อมูล ( isodata ) โดยแยกกลุ่มของพิกเซล
.
( 2 ) มีการจำแนก
แตกต่างจากการจำแนกโดยขาดการควบคุม ดูแลพื้นที่ บนพื้นฐานของความรู้ priori เพื่อแยกแยะความแตกต่างจากประเภทต่างๆของลักษณะพื้นผิวตัวอย่างบนภาพจากระยะไกล ' , then set up templates and take automatic identification. The basic process of image classification as shown below:
การแปล กรุณารอสักครู่..