Finger Flexion Motion Inference from sEMG Signals

This paper provides a method to infer finger flexing motions using a 4-channel surface Electronyogram （sEMG）. Surface EMGs are hannless to the humnan body and easily done. However, they do not reflect the activity of specific nerves or muscles, unlike invasive EMCs. On the other hand, the non-invasive type is difficult to use for discriminating various motions while using only a small number of electrodes. Surface EMG data in this study were obtained from four electodes placed around the forearm. The motions were the flexion of each 5 single fingers （thumb, index finger, middle finger, ring finger, and little fingers）. One subject was trained with these motions and another left was untrained. The maximum likelihood estimation method was used to infer the finger motion. Experimental results have showed that this method could be useful for recognizing finger motions.The average accuracy was as high as 95%.

Modified Laser Speckle Imaging with Adaptive Window

One of the advantages of laser speckle is detecting microvascular through image processing. This paper proposes a new image processing method for laser speckle, adaptive window method that adaptively processes laser speckle images in the space. Disadvantage of conventional fixed window method is that it uses the same window size regardless of target areas. Inherently laser speckle contains undesired noise. Thus a large window is helpful for removing the noise, but it results in low resolution of image. Otherwise a small window may detect micro vascular but it has limits in noise removal. To overcome this trade-off, the concept of adaptive window method is newly introduced to conventional laser speckle image analysis. In addition, the modified adaptive window method applied to other selection images. We have compared conventional Laser Speckle Contrast Analysis （LASCA） and its variants with the proposed method in terms of image quality and processing complexity, Moreover compared the result of the accompamed changing sdection images have also been compared.

Recognition of Finger Motion with sEMG and Gyrosensor Signals

A novel method to infer the finger flexing motions of various arm postures is proposed. From the gyroscope signal, the authors recognized forearm posture using K-means clustering method. Then finger motion inferred. For finger motion inference, Gaussian model of information entropy and maximum likelibood method was utilized. Experimentally it is obtained that the average recognition rate with the forearm posture inference is much higher than those without the inference by 30.7%.