2020 Digital Image Computing: Techniques and Applications (DICTA)
Scene text detection has been gaining a lots of focus in research. Even though the recent methods... more Scene text detection has been gaining a lots of focus in research. Even though the recent methods are able to detect text in complex background having complex shapes with a fairly good accuracy, they still suffer from issues of limited receptive field. These fail from detecting extremely short or long words hence failing in detecting text words precisely in document text images. We propose a new model which we call W-A net, because of it's W shape with the middle branch being Atrous convolutional layers. Our model predicts a segmentation map which divides the image into word and no word regions and also, a boundary map which helps to segregate closer words from each other. We use Atrous convolutions and Deformable convolutional layers to increase the receptive field which helps to detect long words in an image. We treat text detection problem as a single problem irrespective of the background, making our model suitable of detecting text in scene or document images. We present our findings on two scene text datasets and a receipt dataset. Our results show that our method performs better than recent scene text detection methods which perform poorly on document text images, especially receipt images with short words.
In the previous decade, there has been a considerable rise in the usage of smartphones.Due to exo... more In the previous decade, there has been a considerable rise in the usage of smartphones.Due to exorbitant advancement in technology, computational speed and quality of image capturing has increased considerably. With an increase in the need for remote fingerprint verification, smartphones can be used as a powerful alternative for fingerprint authentication instead of conventional optical sensors. In this research, wepropose a technique to capture finger-images from the smartphones and pre-process them in such a way that it can be easily matched with the optical sensor images.Effective finger-image capturing, image enhancement, fingerprint pattern extraction, core point detection and image alignment techniques have been discussed. The proposed approach has been validated on FVC 2004 DB1 & DB2 dataset and the results show the efficacy of the methodology proposed. The method can be deployed for real-time commercial usage.
2020 Digital Image Computing: Techniques and Applications (DICTA)
Scene text detection has been gaining a lots of focus in research. Even though the recent methods... more Scene text detection has been gaining a lots of focus in research. Even though the recent methods are able to detect text in complex background having complex shapes with a fairly good accuracy, they still suffer from issues of limited receptive field. These fail from detecting extremely short or long words hence failing in detecting text words precisely in document text images. We propose a new model which we call W-A net, because of it's W shape with the middle branch being Atrous convolutional layers. Our model predicts a segmentation map which divides the image into word and no word regions and also, a boundary map which helps to segregate closer words from each other. We use Atrous convolutions and Deformable convolutional layers to increase the receptive field which helps to detect long words in an image. We treat text detection problem as a single problem irrespective of the background, making our model suitable of detecting text in scene or document images. We present our findings on two scene text datasets and a receipt dataset. Our results show that our method performs better than recent scene text detection methods which perform poorly on document text images, especially receipt images with short words.
In the previous decade, there has been a considerable rise in the usage of smartphones.Due to exo... more In the previous decade, there has been a considerable rise in the usage of smartphones.Due to exorbitant advancement in technology, computational speed and quality of image capturing has increased considerably. With an increase in the need for remote fingerprint verification, smartphones can be used as a powerful alternative for fingerprint authentication instead of conventional optical sensors. In this research, wepropose a technique to capture finger-images from the smartphones and pre-process them in such a way that it can be easily matched with the optical sensor images.Effective finger-image capturing, image enhancement, fingerprint pattern extraction, core point detection and image alignment techniques have been discussed. The proposed approach has been validated on FVC 2004 DB1 & DB2 dataset and the results show the efficacy of the methodology proposed. The method can be deployed for real-time commercial usage.
Uploads
Papers by Sukhad Anand