Edge detection is an elementary step used in various image processing applications. The main problems in existing edge detection algorithms are poor edge localization, less noise removal capacity, unable to detect edges in complex background images and inability to properly detect the color edges in images. In this paper a sequential hybrid approach is proposed to overcome all the limitations of existing edge detection algorithms. The operations performed by image edge detection algorithm can be computationally expensive and takes lots of execution time for processing the data. This research work also improves hybrid color based image edge detection technique by using the data parallelism approach. The comparison among sequential and parallel edge detection will be drawn based upon different parallel metrics. The experimental results have shown that parallel strategy achieves a performance gain of 68% as compared to sequential approach.