Research on Image Enhancement Algorithm Based on Underwater Target Recognition Accuracy
Aiming at the blurring and color deviation of underwater images caused by underwater light absorption;The problem of decreased accuracy in target recognition due to different poses and mutual occlusion of underwater organisms is addressed by using four methods:RGHS im-age enhancement,IBLA color gamut adjustment,sharpening value adjustment,and contrast pa-rameter to process images in the dataset,in order to restore the original color of the image,re-store image details,and improve recognition accuracy.Experiments have shown that image pro-cessing can effectively improve recognition accuracy.By organically combining four image pro-cessing methods,it has been proven that using multiple image processing methods can improve recognition accuracy.For balanced overall recognition accuracy,using the IBIA color gamut ad-justment is sufficient;Suitable image processing combination schemes are also provided for dif-ferent organisms.
万方数据
维普
