Convolutional neural network method and support vector machine method-based image identification method for imaging department

Abstract

The invention discloses a convolutional neural network method and support vector machine method-based image identification method for an imaging department. The method comprises the following steps ofcollecting an original medical image as a sample, and performing grayscale processing on the original medical image by using a weighted grayscale algorithm to obtain a grayscale image; processing thegrayscale image by using histogram equalization to obtain an equalized grayscale image, performing edge detection on the equalized grayscale image by using an improved Isotropic Sobel edge detectionoperator to obtain an edge image, and performing binarization on the edge image by using an adaptive threshold algorithm to obtain a binarized medical image; processing the binarized medical image byusing morphological arithmetic operation to obtain a medical candidate region image, forming training data by the candidate region image, initializing a convolutional neural network, inputting the training data to the convolutional neural network, and extracting feature data; and transmitting the feature data extracted by training a convolutional neural network model to a support vector machine for performing training, and inputting extracted test feature data to a medical image identification training model for performing judgment, thereby finally obtaining an accurate medical image identification result.

Description

technical field [0001] The invention relates to the field of image processing for medical imaging, in particular to an image recognition method for imaging based on a convolutional neural network method and a support vector machine method. Background technique [0002] At present, many famous scholars at home and abroad are committed to the research of image segmentation algorithms. There are mainly four image segmentation methods: threshold-based, edge detection-based, region-based, and energy-based segmentation. [0003] Threshold-based segmentation methods mainly include histogram concave analysis method, maximum inter-class variance method, threshold interpolation method, etc. This type of method is intuitive, simple and efficient, but due to the complexity of the image, the selection of threshold value has become a major challenge for this type of method; The more classic algorithms in the segmentation method based on edge detection include Sobel, Prewitt, Laplace and C...

AI Technical Summary

Problems solved by technology

[0004] Although there are many image segmentation methods, each has its own disadvantages. Even though the level set method has achieved good results in the field of image segmentation, because the level set method needs to manually initialize the level set function, on the one hand, it increases the amount of labor. On the other hand, different initialization regions lead to great differences in the segmentation results and convergence of the algorithm. The idea of ​​the cloud model effectively integrates the randomness and fuzziness in natural language to form a quantitative and qualitative interface. Therefore, the combination of cloud model and level set method has broad application prospects in the field of image segmentation

However, the current processing and segmentation of medical images has low recognition ability due to its complexity and variability. It does not have a good recognition performance for doctors' examination and treatment, and sometimes some small changes cannot be observed in time. , leading to serious consequences

Images