研究目的
To propose a new developed approach to extract road from high-resolution panchromatic remotely sensed image based on improved ant colony optimization method.
研究成果
The experimental results show that this method can improve the correctness of road extraction on the premise of ensuring the integrity of high-resolution remotely sensed imagery, thus improving the quality of the final extraction road. Therefore, the method proposed in this paper is of great significance in automobile navigation, map automation, new road detection, road disaster prevention, dredging traffic flow and so on.
研究不足
The algorithm does not consider the computation time as the first consideration in the research, which is also the disadvantage of the ant colony optimization algorithm. The computation time will increase rapidly as the number of pixels in the processing image increases.
1：Experimental Design and Method Selection:
The methodology involves improving the deployment of ants and the heuristic function to extract complete road information. Ants are deployed at the edge of the image and move forward with the guidance of the neighbor gray level.
2：Sample Selection and Data Sources:
GF-2 panchromatic remote sensing imagery from Xi'an 503 research Institute was used to detect the road in the Northwest Mountain area.
3：List of Experimental Equipment and Materials:
The tool used in the experiment is eclipse, implemented in the development environment of Isula framework using Java.
4：Experimental Procedures and Operational Workflow:
Ants are placed on the edge of the rectangle imagery and select the next pixel point according to the pheromone value of each pixel and the value of the heuristic function. The path completed from the initial point to the end point is defined as the trajectory of the ant.
5：Data Analysis Methods:
The pheromone matrix is normalized to display more visually, and mathematical morphology processing is used to extract breakpoints and smooth edges in the road.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
