Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

3results about How to "Improve image recognition" patented technology

A limb fixation device for mouse cervical spine imaging

ActiveCN224441512UGuaranteed stabilityMake sure the structural limits are fixedRadiologyCERVICAL SPINE DISEASE
This utility model discloses a limb fixation device for cervical spine imaging in rats, relating to cervical spondylosis research. It includes a fixation platform and a limiting component. The limiting component is installed in the middle of the top surface of the fixation platform, and a label is installed at one corner of the top surface. A supplementary light is located in the middle of one end of the top of the fixation platform, and fixing angle plates are symmetrically arranged on the left and right sides of the fixation platform. The limb fixation device for cervical spine imaging provided by this utility model has several adjustment holes around the placement area in the middle of the operating platform. Combined with the structural design of the limiting component, it can be appropriately adjusted within a certain range according to the rat's body structure, thereby ensuring structural fixation of the rat and guaranteeing the flexibility and convenience of the device. The label is used to help record data labels in the rat's images, improving image recognition.
Owner:WANGJING HOSPITAL OF CHINA ACAD OF CHINESE MEDICAL SCI

Portable X-ray imaging matching device for rapidly screening fish bone malformation

The invention relates to the technical field of X-ray imaging devices, in particular to a portable X-ray imaging matching device for quickly screening fish bone deformity, which comprises a fish body fixing and guiding module, a radiation source positioning module, a bone alignment imaging module, a gray scale fracture detection module and a deformity screening conclusion output module. According to the method, the influence of image interference information is reduced by comparing the position relation between the fitting point and the guiding track, pushing the tail of the fish body to be aligned with the guiding center line, continuously monitoring the offset position of the intersection point, correcting the corresponding relation between the irradiation path and the axis of the fish body, extracting the skeleton trunk area and removing edge fracture pixels; analyzing transverse contraction and included angle deviation trends in the image, delimiting abnormal position sections in the skeleton structure, highlighting the overall extension form and structural continuous trend of the skeleton contour in the image, and matching fish body individual differences and imaging environment changes to obtain a fish body image; and the image identification degree, the abnormal response speed and the result judgment stability degree in the malformation screening process are improved.
Owner:HUBEI UNIV +1

A target detection method based on difference feature compensation under aggregated multi-scale information

ActiveCN115984660BImprove image recognitionInstrumentsEngineeringComputer vision
The application provides a target detection method based on differential feature compensation under polymeric multi-scale information, comprising constructing a multi-scale feature fusion module, selecting different receptive field convolution and various pooling methods to process input features to obtain differential features containing multi-scale information; constructing a parallel attention module to calculate spatial and channel attention maps in parallel, wherein various pooling methods are jointly used when inferring the attention map; combining the constructed multi-scale feature fusion module and parallel attention module in the best strategy to form a differential attention module for a feedforward convolutional neural network, and embedding the differential attention module into the network to obtain optimal performance. The application proposes a differential attention module (DAM) for a feedforward convolutional neural network, adopts a unique differential structure, and completes feature information compensation in the global through multi-scale feature fusion and attention mechanism to enhance the representation ability of features.
Owner:DALIAN POLYTECHNIC UNIVERSITY