Z. Jijun et alii, Frattura ed Integrità Strutturale, 34 (2015) 590-598; DOI: 10.3221/IGF-ESIS.34.65 590 Defect detection of wire rope for oil well based on adaptive angle Zhang Jijun, Meng Xiangqing Southwest Petroleum University, China , A BSTRACT . This paper uses a digital image processing method which is based on texture feature of steel cable to detect the fracture of steel wire. At first, it uses a modified homomorphic filtering method to eliminate environment heterogeneous shining. Next, it obtains the body image of the steel line by using the method of edge detecting and section counting filtering to detect the bunch part of steel wire. By using an improved Radon transformation method to indicate if those steel wire are in good condition or not. Finally, by using BP neural network model it aims to judge the final result. Test result shows that this method is easy to use and fulfill real time request. K EYWORDS . Homomorphic filtering method; Edge detection; Integral projection; BP neural network. I NTRODUCTION s a flexible component, wire rope has been widely used in engineering. The common cableway, crane, hoister and elevator have all used wire rope, it occupies an important position in industrial sectors such as metallurgy, machinery, chemical industry, oil, transportation, tourism and construction industry, etc., therefore, it directly affects people’s daily life, even the development of the national economy [1]. Wire rope is high in strength, good in tenacity, as well as strong in bearing capacity, which makes low noise and exert stability during work. However, as a bearing component in engineering, it will deform, be corroded, fracture or even suddenly break down due to the load and tensile force it bears for a long time. It affects the normal operation of the equipment, even threaten the life security of the public, Great importance should be attached to its user’s safety. In the drilling platform of oil and gas in this paper, wire rope is supervised through naked-eye identification and regular replacement in order to extend the service life of wire rope and guarantee the safety of its application. However, erroneous judgment and negligence are inevitable in the check by naked eyes, and the regular replacement will also result in wasting, consuming unnecessary man power and material resource. Hence, from the view of both economic benefit and social benefit, developing a new method to defect detection of wire rope is an inexorable trend [2-3]. In order to increase the degree of intelligentization of the nondestructive testing of wire rope, improve the objective & unscientific parts and low efficiency and provide more detailed and perfect information of fracture of wire, and evaluate the damage, this paper discusses a way of defect detection of wire rope for oil well based on adaptive angle by regarding the projection features of the images of wire rope as the basis of defect identification. A