Canny Edge Detection Analysis Based on Parallel Algorithm, Constructed Complexity Scale and CUDA

Abstract

Edge detection is especially important for computer vision and generally for image processing and visual recognition. On the other hand, digital image processing is widely used in multiple science fields such as medicine, X-ray analysis, magnetic resonance tomography, computed tomography, and cosmology, i.e. information collection from satellites, its transferring, and analysis. Any step of image processing, from obtaining the image to its segmentation and object recognition is followed by image noise. The processing speed is vital in popular fields that demand image analysis in real time. In this work, we have proposed an approach of parallel computing of the Canny algorithm using CUDA technology, the complexity of object recognition is analyzed according to the type of the image noise and the level of its density. The sequenced implementation on GPU and the parallel implementation on GPU was considered. The results were analyzed for efficiency and reliability. Also, parallel acceleration is calculated according to the size of the incoming image. The manipulations with the image showed the growth of processing speed of 68 times, whereas the manipulations with the size of the kernel showed the growth of processing speed of 26 times. Another contribution of this work is the analysis of the complexity of object recognition depending on the type of image noise and the level of its density. Furthermore, the increase of Gaussian noise density linearly increases the complexity of object recognition.

Enhancing quality measurement for visible and invisible watermarking based on M-SVD and DCT

“…Salt and pepper noise denotes a variety of digital image disturbances characterized by arbitrary distribution of black and white pixels dispersed across an image's expanse, resembling grains of salt and pepper [35]. To describe the presence of salt and pepper noise mathematically, we can use (4).…”

A parallel Canny edge detection algorithm based on OpenCL acceleration

“…The time-consuming trend of the OCL_Canny parallel algorithm in Fig 11 is basically consistent with the situation reflected in its time complexity. Literature [15] and literature [17] reported the implementation results of the Canny algorithm under CUDA architecture, literature [16] reported the implementation results of the Canny algorithm under OpenCL architecture, and literature [11] reported the implementation results of the Canny algorithm under FPGA computing architecture. The data shown in these literatures are compared with the time-consuming of the algorithms in this paper, as shown in Table 7.…”

“…The data shown in these literatures are compared with the time-consuming of the algorithms in this paper, as shown in Table 7. According to Table 7, the Canny serial algorithm implemented in references [11,15,17] takes more time to run on the six groups of images involved than the CPU_Canny algorithm in this paper. The four parallel algorithms in literature [11], literature [15], literature [16], and literature [17] all consume significantly more time than the CUDA_Canny and OCL_Canny parallel algorithms in this paper on the 10 sets of images involved.…”

“…Jin et al [16] proposed a Canny edge detection algorithm under OpenCL architecture, which achieves 1.42 times speedup at 2048 × 1536 gray-scale image resolution without considering data transmission. Mochurad [17] proposed a parallel computing method of the Canny algorithm using CUDA technology. The performance of the algorithm for processing gray-scale images with image frames of 10240 × 10240 is improved by 68 times.…”

The Markov Concept of the Energy Efficiency Assessment of the Edge Computing Infrastructure Peripheral Server Functioning over Time

“…In such a case, the researcher's task is to find an appropriate control scheme for the components of the simulated computing device or their associated composition by analyzing the corresponding characteristics of the queuing system under certain initial conditions. Among the numerous scientific and applied works on this topic, it is worth noting the works [17][18][19][20][21][22] (systems with D policy), [19,[23][24][25] (systems with N policy), [24][25][26][27][28] (systems with T policy). Despite the sufficient development of the subject area, research in this direction is constantly relevant.…”