Air-jet texturing, also known as cold fluid texturing, is one of the three major processes used to impart bulk to flat continuous filament yarns via a supersonic turbulent air flow. Air-jet texturing produces loops which surround the entangled, relatively straight parts of the filaments in the core of the textured yarn. Several classical techniques as well as more modern image analysis based methods have been used to characterise the air-jet textured yarn structure via parameters such as core diameter, overall diameter, average loop size and loop frequency. However, all these techniques use a single image which is not representative of the three-dimensional structure. Research is reported to develop a more comprehensive technique, capable of characterising the structure via Canny edge detector and segmentation techniques. The proposed image analysisbased technique uses two images captured at right angles from the same part of a yarn. Validation against manual measurements suggests that the technique has good accuracy and potential to characterise air-jet textured yarn structures.