ABSTRACT:

The confidence of decisions taken with image data are dependent on the final output of the processed image. The choice of software to be used by an operator for image processing ideally should depend on the purpose of the task at hand which in turn would depend on the capabilities of each of the software. However, choice of a software for image processing as commonly practiced by operators oftentimes depended on the operator's proficiency in a particular software. This has resulted in a variety of results whose interpretation could impair accurate decision making. The aim of this research is to carry out a comparative study on the performance of four image processing software packages which includes ENVI, IDRISI, ILWIS and MATLAB. A LANDSAT Imagery, Path 191 and Row 55 acquired December, 2002 was used in this study. Six criteria were used to test the performance of each software. In addition, classification and post classification accuracy assessment of the processed image from each software was carried out. The result shows that ENVI has the least launch time and easiest in conversion to software specific format, signature training and image classifications. ILWIS is the best in colour composite generation while IDRISI gives the best post-classification results