Mechanical Characteristics Extraction Of Cardiac And Gesture Recognition Based On Improved ASM

Cardiovascular disease is the primary cause of death in industrialized countries,and it is always harmful to mankind's health.Currently,medical ultrasonic imaging is widely used in the evaluation of cardiac function and heart disease diagnosis because of its advantages such as non-invasive,real-time,cheap,good repeatability,high sensitivity,etc.However,due to the special ultrasonic imaging principle,inevitably there are a lot of speckle noise in ultrasonic images,so the segmentation result can't meet clinical requirements on the stability,robustness,and speed,and there are a lot of pseudo-boundary,discontinuous boundary,it is difficult to promote segmentation method.Active Shape Model(ASM)is a kind of image segmentation method based on statistical analysis,which iteratively adjust model to fit the shape of target boundary in the sample deformation mode,and which is widely used in face recognition.As a result,the accuracy of heart shape model built is related to the segmentation result,and different training and alignment method will affect the shape model.For this situation,the training and alignment of ASM are improved in this paper,and then the improved shape model which is applied to the echocardiographic images is put forward.Segment cardiac ultrasonic image sequences of 4?5 cardiac cycles of the patient to get cross section area of four cavities,and then fit area and analyze mechanical characteristics of atrial and ventricular.The experimental results prove that the improved shape model has a better segmentation effect,in order to know the truth of heart disease and make the correct diagnosis for doctors,so the image segmentation result must be accurate.Along with China entering the aging society,old people gradually become the center of the family life.People are too busy to care the old people's daily life because of all kinds of life and work pressure.We can install a camera to monitor the old people's life to solve this problem,they can pose a gesture in front of the camera if the old man feel uncomfortable or fall,gesture information lead to alarm,avoid to hurt the old man's health when they stay at home alone.The improved ASM is also applied to the area of gesture recognition in this paper.First,collect different people's gestures and then establish gesture shape model,finally validate recognition effect of every gesture.The experimental result shows that recognition rate is high and the effect is ideal if the background is simple.