Transport affects agricultural marketing because it is the only means by which farmers can transport their produce to the market. Reliable road path is essential for transporting the cultivated crops to the market at right time. Heavy traffic, railway crossings, road repair and unexpected road blocks increases the transportation time. The increase in transportation time results in loses of crop freshness and quality which reduces the profit for farmers. The impacts of road infrastructure on agricultural output and productivity are particularly important in rural development. In this paper, we propose a methodology to detect reliable road path with reduced transportation time in the presence of road blocks or damages. The road network with the length and width of roads (both damaged as well as good) can be extracted accurately from a given remote sensing image. Discontinuities of roads are identified (due to damage or block) using multilevel searching by iteratively increasing the jump step size in pixels. The experimental results shows that the efficiency and accuracy of the proposed methodology. © 2015 IEEE.