基于监测特殊机器人图像识别及射击准确度装置

Image Recognition and Strike Accuracy Device Based on Monitoring Special Robot

为解决机器人无法对连续不断改变的动态物体进行快速识别以及准确找到图像中心位置的问题,设计一种训练特殊机器人图像识别和监测射击准确度并进行实时报靶的装置。通过Mini PC输出和更换图像,讨论一种基于有限脉冲响应的多项式预测滤波器和改进的中值滤波相结合的混合滤波方法,对压力传感器的输出信号进行滤波降噪,STM32通过识别消噪和放大后的信号判断受击打的位置;并通过数码管进行显示。实验结果表明,该装置具有较高的灵敏度和准确性。

To solve the problem that the robot can not quickly identify continuously changing dynamic objects and accurately find the center position of the image,a device is designed to train special robot image recognition,monitor shooting accuracy and report target in real time.First,output and replace the image with the Mini PC.Next,pressure sensor output signal polluted by additive noise are processed by mixed filter with polynomial prediction filter of finite impulse response and the improved median filter.Then,the STM32 determines the location of the hit by identifying the de-noised and amplified signals.Finally,the hit data is displayed by digital tube.The experimental results show that the device has high sensitivity and accuracy.