Research On The Algorithm Of Arm Vein Image Perception And Needle Point Location

Venipuncture is a common diagnosis and treatment method in clinical medicine.Hundreds of millions of patients all over the world need venipuncture treatment every day.However,because the puncture process relies heavily on the naked eye and experience of medical staff,some inexperienced medical staff have a higher puncture failure rate when facing patients with unobvious vein characteristics such as obesity,infants and young children.Repeated venipuncture procedures not only increase the patient's pain,but more dangerously,secondary infections may occur.In addition to puncture work,medical staff also undertake a large number of other complicated tasks.It often happens that patients queue up for venipuncture.Therefore,it is particularly important to replace medical staff with puncture robots.This thesis analyzes and compares the vein sensing technology on the market,and can be transplanted into the existing automatic blood sampling robot platform,to determine the image and ultrasound-guided vein sensing and positioning technology.Use the principle of near-infrared imaging to complete the hardware selection of the sensing module and the design of the image acquisition scheme.Based on the classical algorithm Contrast Limited Adaptive Histogram Equalization(CLAHE),a vein image enhancement algorithm based on the fusion of multi-directional multi-scale Gabor filtering and CLAHE algorithm is proposed to improve image contrast.Then refine the image to obtain the vein skeleton,traverse the intersection of each curve on the basis of the skeleton,select the appropriate needle point position,and calculate the target position of the needle tip of the end effector of the blood sampling machine through coordinate transformation.After experiments,the most suitable lighting conditions have been tested to improve the image quality of the camera.Through experimental verification,the algorithm can complete the perception and positioning of images in different images.The algorithm can be transplanted to the existing automatic blood collection machine in the laboratory to complete the overall puncture blood collection movement.The algorithm proposed in this study can well perceive the position of the venous blood vessel and locate its needle point posture.It can be applied to some puncture blood collection robots,which can effectively improve the success rate of vein access and provide colleagues with a way of solving problems.And method.