|Title||A fast method to determine co-occurrence texture features using a linked list implementation|
|Publication Type||Conference Paper|
|Year of Publication||1996|
|Authors||Clausi, D. A.|
|Conference Name||26th International Symposium on Remote Sensing of Environment and 18th Annual Symposium of the Canadian Remote Sensing Society|
|Conference Location||Vancouver, BC, Canada|
A linked list approach has been developed to efficiently calculate texture features based on cooccurrence probabilities. The commonly used matrix based approach (the grey level cooccurrence matrix or GLCM) requires an unreasonable amount of computation, especially for image segmentation purposes. The linked list approach calculates exactly the same results while significantly decreasing the time to both generate the cooccurrence data and calculate the texture features. The full dynamic range may be maintained without the dramatic increase in computation time that would be experienced by the GLCM approach, however, the behaviour of the statistics changes with different grey level quantizations. This paper describes the implementation of a linked list algorithm, demonstrates its applicability, and investigates the validity of the cooccurrence texture features.