A new approach for tuning the parameters of MultiScale Retinex (MSR) based color image enhancement algorithm using a popular optimization method, namely, Particle Swarm Optimization (PSO) is presented in this paper. The image enhancement using MSR scheme heavily depends on parameters such as Gaussian surround space constant, number of scales, gain and offset etc. Selection of these parameters, empirically and its application to MSR scheme to produce inevitable results are the major blemishes. The method presented here results in huge savings of computation time as well as improvement in the visual quality of an image, since the PSO exploited maximizes the MSR parameters. The objective of PSO is to validate the visual quality of the enhanced image iteratively using an effective objective criterion based on entropy and edge information of an image. The PSO method of parameter optimization of MSR scheme achieves a very good quality of reconstructed images, far better than that possible with the other existing methods. Finally, the quality of the enhanced color images obtained by the proposed method are evaluated using novel metric, namely, Wavelet Energy (WE). The experimental results presented show that color images enhanced using the proposed scheme are clearer, more vivid and efficient.
This paper presents the development of a new algorithm for Gaussian based color image enhancement system. The algorithm has been designed into architecture suitable for FPGA/ASIC implementation. The color image enhancement is achieved by first convolving an original image with a Gaussian kernel since Gaussian distribution is a point spread function which smoothen the image. Further, logarithm-domain processing and gain/offset corrections are employed in order to enhance and translate pixels into the display range of 0 to 255. The proposed algorithm not only provides better dynamic range compression and color rendition effect but also achieves color constancy in an image. The design exploits high degrees of pipelining and parallel processing to achieve real time performance. The design has been realized by RTL compliant Verilog coding and fits into a single FPGA with a gate count utilization of 321,804. The proposed method is implemented using Xilinx Virtex-II Pro XC2VP40-7FF1148 FPGA device and is capable of processing high resolution color motion pictures of sizes of up to 1600x1200 pixels at the real time video rate of 116 frames per second. This shows that the proposed design would work for not only still images but also for high resolution video sequences.
Object recognition in the video sequence or images is one of the sub-field of computer vision. Moving object recognition from a video sequence is an appealing topic with applications in various areas such as airport safety, intrusion surveillance, video monitoring, intelligent highway, etc. Moving object recognition is the most challenging task in intelligent video surveillance system. In this regard, many techniques have been proposed based on different methods. Despite of its importance, moving object recognition in complex environments is still far from being completely solved for low resolution videos, foggy videos, and also dim video sequences. All in all, these make it necessary to develop exceedingly robust techniques. This paper introduces multiple moving object recognition in the video sequence based on LoG Gabor-PCA approach and Angle based distance Similarity measures techniques used to recognize the object as a human, vehicle etc. Number of experiments are conducted for indoor and outdoor video sequences of standard datasets and also our own collection of video sequences comprising of partial night vision video sequences. Experimental results show that our proposed approach achieves an excellent recognition rate. Results obtained are satisfactory and competent.
This paper presents a new color image enhancement technique based on modified MultiScale Retinex(MSR) algorithm and visual quality of the enhanced images are evaluated using a new metric, namely, wavelet energy. The color image enhancement is achieved by down sampling the value component of HSV color space converted image into three scales (normal, medium and fine) following the contrast stretching operation. These down sampled value components are enhanced using the MSR algorithm. The value component is reconstructed by averaging each pixels of the lower scale image with that of the upper scale image subsequent to up sampling the lower scale image. This process replaces dark pixel by the average pixels of both the lower scale and upper scale, while retaining the bright pixels. The quality of the reconstructed images in the proposed method is found to be good and far better then the other researchers method. The performance of the proposed scheme is evaluated using new wavelet domain based assessment criterion, referred as wavelet energy. This scheme computes the energy of both original and enhanced image in wavelet domain. The number of edge details as well as wavelet energy is less in a poor quality image compared with naturally enhanced image. Experimental results presented confirms that the proposed wavelet energy based color image quality assessment technique efficiently characterizes both the local and global details of enhanced image.