Deteksi keretakan permukaan perkerasan lentur jalan raya (Studi kasus: tanah lunak di Banjarmasin)

Abstract

One type of soft soils is peat, which physically and technically lacks the requirements and conditions for road construction. The reason is that peat contains a very high water content and compressibility, and has a low soil bearing capacity. The total length of the roads in the city of Banjarmasin is 577.11 km, which consists of 1.5% of state roads, 3.8% of provincial roads and 95.3% of municipality roads. Of that amount, asphalt pavement covers length of 465.63 km (80%) and the rests are gravel/stone pavements. Of that length, the defect conditions can be categorized as medium damaged of 169.54 km, damaged of 53.93 km, and heavily damaged of 53.92 km. The low bearing capacity of subsoil, which is formed by peat dominates the major cause of road cracking in Banjarmasin. Therefore, road quality evaluation needs to be conducted frequently so that a rehabilitation can be carried out and the road service performance can be maintained. It is common in Indonesia that the evaluation of damaged road surfaces still relies on manpower (manual) and it is done visually. Meanwhile in other countries, an automatic system for detecting road defects has been implemented, which is found to be more time- and cost-efficient and not endangering the workers. Recent researches have demonstrated image processing approach for detecting road defects. This research proposes an intelligent road evaluation system by applying neural network algorithms. The method is based on digital image processing gained from camera to detect road surface cracks and damages