Pacific Oyster Gonad Identification and Grayscale Calculation Based on Unapparent Object Detection

The plumpness of the Pacific oyster gonad, the reproductive organs of both male and female oysters which are buried within the flesh of the oyster in the shell, has important implications for the quality and breeding of subsequent parents. At present, only the conventional method of breaking their shells allows for the observation and study of the interior tissues of Pacific oysters. In this paper, the gonad of Pacific oyster was observed by small animal Magnetic Resonance Imaging (MRI), and a multi-effective feature fusion network algorithm R-SINet was proposed for the detection of unapparent target, in Nuclear Magnetic Resonance (NMR) images, which can effectively solve the problem that the gonads of Pacific oysters are difficult to identify from the background images. In addition, the gray histogram of the segmented gonad region was calculated, and it was found that the female and male had differences in gray value. The sex of oyster was nondestructively detecting by this task. Firstly, established the Oyster gonad datasets; secondly, a compact pyramid refinement module that combines with high-level semantic features and low-level semantic features was proposed, designed a lightweight decoder to improve the accuracy of feature fusion; thirdly, a switchable excitation model capable of adaptive recalibration is proposed to obtain an attention map. Experimental results on the Oyster gonad datasets demonstrate the effectiveness of the method. Comparing R-SINet’s experimental findings to those of popular algorithm models, such as the benchmark algorithm SINet_v2, revealed promising results.