Our challenge is detection of anti-personnel landmines from images captured in the visible spectrum where the landmine is partly covered by soil, vegetation and clutter. This paper provides five different texture-based classical image segmentation techniques viz. Symlet Wavelet-based, Phase Stretch Transform-based, Fractal Texture Analysis-based, Graph-based, and Active Contour-based segmentation techniques for extracting maximum portion of the existing landmine. The Active Contour-based segmentation technique provided the best overall results showing the detected anti-personnel landmine in a single gray level intensity separate from the background (in a different gray level intensity) for all the images. The output images from the Active Contour-based technique have also been subjected to Coincidence Measure for the evaluation of the segmentation process. This clearly proves the validity and suitability of this technique. The current study provides the foundation for further studies towards error-less detection of anti-personnel landmines.