Conventional edge detection involves convolving images with the kernels as the 1st step, then followed by local orientation calculation, if required and thresholding to detect the edges. Where ever necessary the algorithm would also carry out some kind of image modification similar to pre-processing to enhance the image quality for better edge detection. In general, convolution is the preliminary and the most crucial step because the convolved images obtained may or may not retain all the details present in the original image. Abs-Laplacian is a newer technique and has been demonstrated to be better in comparison to Sobel and Prewitt. Here in the paper, I present recent developments that have been undertaken so far in terms of the advantage we have with the abs-Laplacian along with a comparative analysis carried over a set of 9 kernels/edge detection techniques. Edge quality determined by abs-Laplacian is clear and comparable to the best available technique. In addition the kernel requires a complexity of only 7 additions against 15 and 13 respectively for Sobel and Prewitt. While the Robert’s cross operator performs the faster with a mere 3 additions, the previous and also the current analysis would show that the edge quality does not fare better in comparison to newer technique. In order to demonstrate the potential abilities of the technique to resolve fine details in the image, the edge detection was carried out on high-resolution complex images so that we can compare the edges together with its original image. I would like to conclude here saying that technique offer best in terms of edge quality and also in speed for real time imaging applications.