In order to reduce the influence of noise in the detection process of ecological environment monitoring images,so as to obtain more accurate and available information,in this paper,we take ecological environment monitoring images as the research object and conduct a systematic study on the detection methods of multi-feature parameters.The correlation between the ecological environment monitoring image and the gray value of neighboring pixels is used to locate the source of noise pollution.Furthermore,the location result of the noise pollution source is used to mark the corresponding matrix elements,and the degree of pixel noise pollution is judged.Then,the processed filtering result is used as the image gray value,and the gray co-occurrence matrix is used to select characteristic parameters such as energy,contrast and entropy with strong description performance to extract the characteristics of the ecological environment monitoring image.Finally,the image feature components of the neighborhood consistency and directionality measurement are used to realize the ecological environment monitoring image detection.The ecological environment monitoring images with different signal-to-noise ratio and salt and pepper noise density were tested respectively.Experiments show that the proposed method can accurately extract image details in noise.Therefore,this method has better noise suppression performance and inspection-free performance,and its practical applicability is high.
展开▼
