Parallel edge detection based on digital differentiator approximation

Edge detection is an established image processing field with various applications. The work uses the novel image edge detection technique based on the Al-Alaoui FIR Filter for different purposes that include sharpening, histogram equalization, and range filtering. The edge detection technique which is based on applying masks or kernels on images is further mapped to an improved parallel implementation on CUDA using shared memory and coalescence that is compared and proven better than the Image Processing Filter provided through the NVIDIA GPU Computing SDK on various GPUs. This technique is necessary in this field and can be used for any image processing application which involves masks.