Infrared Image Enhancement Algorithm Based On Salient Region Detection And FPGA Implementation

With the development of infrared technology,infrared imaging system is widely used in military,civilian and other fields.However,due to the limitation of the infrared detector technology and the influence of external environment,the infrared image has some characteristics such as low contrast and bad imaging quality,which is not suitable for person eye observation.In order to solve the above problems,an image enhancement system is designed and built on the FPGA platform by using the infrared image enhancement algorithm based on salient regions in this dissertation.First of all,the dissertation studies the development of infrared image enhancement algorithms and their application in FPGA platform at home and abroad.At present,the enhanced algorithms implemented on the traditional hardware platform can not differentiate the target from the background,which makes it difficult to distinguish between the observed target and the background information,and some target information is overwhelmed by too much background information.In order to solve these problems,this dissertation focus on an infrared image enhancement algorithm based on salient region detection.Firstly,the algorithm detects the salient region of the image,and then calculates the difference information of the salient region by layered difference representation.The structural relationship between layers is considered and the high level differential variables are converted to lower levels.The difference vectors of the salient region are calculated by solving the linear equations,and the transformation functions are obtained by connecting the non-salient region difference vectors.The enhanced image is getted by mapping and reconstructing the original image finally.Simulation experiments compares the performance of the proposed algorithm with several common enhancement algorithms from subjective visual and objective indexes.The results show that the proposed algorithm is superior in local contrast and edge retention,and human visual effect is better.Finally,the dissertation designs and builds a real-time image processing system upon the Xilinx Spartan-6 FPGA chip.Simultaneously,the infrared image enhancement algorithm based on salient region detection is verified and implemented in this system.Experiments in different scenes show that the output of the hardware system is basically consistent with the simulation results,and the system has strong adaptability to different scenes and can effectively improve the visual effects of the images.