Extreme Intelligent for Laparoscopic Surgery on Medical Deconvolution using Platform Prediction together with Artificial Autofocusing

Abstract

This research proposes a new technique for Laparoscopic autofocus as follows: (1) we propose blur area detection in image on Laparoscopic Surgery (LS) by using discrete-time Fourier series together with artificial intelligent recognize in blur image. (2) Plate Resorution Sector Coordinate (PRSC) technique and pattern recognition processing are used for searching blur area. (3) The image enhancement is adjusted by artificial automatic-focus with Kernel Platform Prediction algorithm. This processing is more quickly than hardware so that it is intelligent to solve the image problem in real-time. If system can not found any blur areas the adaptive status is displayed in monitor immediately. (4) This process can be used with any kind of moicroscope. We are study performance (2D)²PCA algorithms for recognition in detect blurring area while move Laparoscopic Surgery . One hundred blur images from Laparoscopic Surgery camera and actualize image enhancement with artificial automatic-focus was used for this experiment. The results shown the accuracy of blur detection 99.67 percentage. The (2D)²PCA algorithm is faster for recognize and detect blur area in real-time. The automatic focus process uses Kernel Platform Prediction algorithm together with Multiscale Pointwise Product has more efficiency than other methods. We use Gaussian Fitting for calculate for short range in adjust by hardware pass the circuit. Moreover, the adjustable by manual hardware that slow and unstable is decreased. The processing’s speed of this method is faster than the conservative method (manual adjust) 75.33 percentage base on physical condition of patient. This experimental is working base on Quality Medical Standards (QMS) of Faculty of Medical, Canterbury University.