Wyniki wyszukiwania - BazTech

1

Bayesian network aided grasp and grip efficiency estimation using a smart data glove for post-stroke diagnosis

Dutta D., Modak S., Kumar A., Roychowdhury J., Mandal S.

Biocybernetics and Biomedical Engineering

|

2017

|

Vol. 37, no. 1

44--58

EN

Stroke is one of the major causes behind the increased mortality rate throughout the world and disability among the survivors. Such disabilities include several grasp and grip related impairment in daily activities like holding a glass of water, counting currency notes, producing correct signature in bank, etc., that seek serious attention. Present therapeutic facilities, being expensive and time-consuming, fail to cater the poverty stricken rural class of the society. In this paper, on the basis of an investigation, we developed a smart data glove based diagnostic device for better treatment of such patients by providing timely estimation of their grasp quality. Data collected from a VMG30 motion capture glove for six patients who survived stroke and two other healthy subjects was fused with suitable hypothesis obtained from a domain expert to reflect the required outcome on a Bayesian network. The end result could be made available to a doctor at a remote location through a smart phone for further advice or treatment. Results obtained clearly distinguished a patient from a healthy subject along with supporting estimates to study and compare different grasping gestures. The improvement in mobility could be assessed after physiotherapeutic treatments using the proposed method.

2

Map image binarization and stitching using extraction of regions

Mandal S., Biswas S., Das A. K, Chanda B.

Journal of Theoretical and Applied Computer Science

|

2015

|

Vol. 9, nr 1

28--40

EN

Binarisation methods already reported are inadequate for binarisation of complex documents such as maps due to large intensity variations across the regions and entangled texts with lines representing borders, rivers, roads and similar other components. This paper proposes a new binarisation technique for coloured land map images by extracting the regions and analysing the hue, saturation spread and within class kurtosis. This is a region-wise adaptive algorithm to cope up with the sharp changes of the discriminating features across different regions. Here, local regions are selected as clusters having the same hue and saturation. The regions are individually binarised using the spread of their degree of within class kurtosis and finally combined together. The regions extracted are further utilised for stitching of map documents which contain some portion in common. We use a simple greedy technique using correlation matching to join two or more map images such that information from both can be viewed and compared. Our experiments include 446 colour maps from the map image database created for this purpose and made freely available at website. This work is an extended version of our previous work on map image binarisation [1].

3

Land Map Image Dataset: Ground-Truth And Classification Using Visual And Textural Features

Mandal S., Biswas S., Kumar-Das A., Chanda B.

Image Processing & Communications

|

2014

|

Vol. 19, no. 4

37--55

EN

Research on document image analysis is actively pursued in the last few decades and services like OCR, vectorization of drawings/graphics and various types of form processing are very common. Handwritten documents, old historical documents and documents captured through camera are now being the subjects of active research. However, another very important type of paper document, namely the map document image processing research suffers due to the inherent complexities of the map document and also for nonavailability of benchmark public data-sets. This paper presents a new data-set, namely, the Land Map Image Database (LMIDb) that consists of a variety of land maps images (446 images at present and growing; scanned at 200/300 dpi in TIF format) and the corresponding ground-truth. Using semiautomatic tools non-text part of the images are deleted and the text-only ground-truth is also kept in the database. This paper also presents a classification strategy for map images using which the maps in the database are automatically classified into Political (Po), Physical (Ph), Resource (R) and Topographic (T) maps. The automatic classification of maps help indexing of the images in LMIDb for archival and easy retrieval of the right maps to get the appropriate geographical information. Classification accuracy is also tested on the proposed data-set and the result is encouraging.