Youssef Zaz

Enhanced EPR data protection using cryptography and digital watermarking

In this paper, we propose a novel method to embed Electronic Patient Records (EPR) data in medical images. Indeed, after liberating a zone by compressing the image Least Significant Bit plan using the Huffman coding, the EPR is encrypted by an Elliptic Curve Cryptosystem (ECC) and inserted into this zone. The proposed method improves medical security performance and reduces the computation cost related to data encryption and decryption.

Indoor localization system benchmark based on wireless local network technologies

Localization need is appeared with the evolution of mobile communication devices. Nowadays, smart phones are equipped with many sensors that allow position estimation. Based on localisation, many applications and values added services are developed to meet the increasing users need. In order to design localization system, it requires first to understand the fundamental localization principles and to identify existing methods. A compromise between cost, complexity and performance should be taken into account while choosing the best solution. This paper makes benchmark of existing solutions for RF indoor localization with a detailed presentation of performance and limitations related to the most promising solutions. It presents also the latest researches and stimulates new ways to develop new research in this field in particular for fingperpinting based solution and CSI used metric for WLAN network.

Remote QR code recognition based on HOG and SVM classifiers

QR codes have become useful and efficient data storage tools which are exploited in many commercial applications including product tracking, website redirection, etc. A QR code is a 2-dimensional barcode localised through three finder patterns (three squares characterised by a series of alternative black and white modules at ratios 1∶1∶3∶1∶1) placed in its three corners. QR codes are generally placed in different environments with complex backgrounds (overlapping text, pictures, etc.), and are often captured under unfavourable conditions such as poor lighting. These factors can significantly affect the recognition ability and thus may hinder correct QR code localisation and identification. In order to appropriately address these issues, in this paper, we present a QR code recognition algorithm based on histogram of oriented gradients (HOG) features combined with support vector machine (SVM) classifiers. Using HOG, we extract gradient features of each extracted pattern. Subsequently, the obtained features are passed to two linear SVM classifiers, one trained with finder patterns and one trained with alignment patterns, to remove irrelevant patterns. QR codes are then conveniently localised according to a pattern closeness constraint. In the last stage, the captured code is enhanced by applying a perspective correction followed by image binarisation and morphological processing. Finally, the patterns are decoded using an accurate 2-d barcode decoder. Our proposed approach is designed for an embedded systems using a Raspberry Pi equipped with a HD camera and a small robot carrying the equipment.

QR Code Patterns Localization based on Hu Invariant Moments

The widespread utilization of QR code and its coincidence with the swift growth of e-commerce transactions have imposed the computer vision researchers to continuously devise a variety of QR code recognition algorithms. The latter performances are generally limited due to two main factors. Firstly, most of them are computationally expensive because of the implemented feature descriptor complexities. Secondly, the evoked algorithms are often sensitive to pattern geometric deformations. In this paper a robust approach is proposed, in which the architecture is based on three distinct treatments among others: 1) An image quality assessment stage which evaluates the quality of the captured image in consideration that the presence of blur decreases significantly the recognition accuracy. 2) This stage is followed by an image segmentation based on an achromatic filter through which only the regions of interest are highlighted and consequently the execution time is reduced. 3) Finally, the Hu invariant moments technique is used as feature descriptor permitting removing false positives. This technique is implemented to filter out the set of extracted candidate QR code patterns, which have been roughly extracted by a scanning process. The Hu moments descriptor is able to recognize patterns independently of the geometric transformations they undergo. The experiments show that the incorporation of the aforementioned three stages enhances significantly the recognition accuracy along with a notable diminution of processing time. This makes the proposed approach adapted to embedded systems and devices with limited performances.

QR Code Recognition based on Principal Components Analysis Method

QR (Quick Response) code recognition systems (based on computer vision) have always been challenging to be accurately devised due to two main constraints: (1) QR code recognition system must be able to localize QR codes from an acquired image even in case of unfavorable conditions (illumination variations, perspective distortions) and (2) The system must be adapted to embedded system platforms in terms of processing complexity and resources requirement. Most of the earlier proposed QR code recognition systems implemented complex feature descriptors such as (Harris features, Hough transform which aim at extracting QR code pattern features and subsequently estimating their positions. This process is reinforced by pattern classifiers e.g. (Random forests, SVM) which are used to remove false detected patterns. Those approaches are very computationally expensive. Thus, they are not able to be run in real-time systems. In this paper, a streamlined QR code recognition approach is proposed to be efficiently operable in systems characterized by a limited performance. The evoked approach is conducted as follows: the captured image is segmented in order to reduce searching space and extract the regions of interest. Afterwards a horizontal and vertical scans are performed to localize preliminarily QR code patterns, followed by Principal Component Analysis (PCA) method which allows removing false positives. Thereafter, the remaining patterns are assembled according to a constraint so as to localize the corresponding QR codes. Experimental results show that the incorporation of PCA decreases notably the processing time and increase QR code recognition accuracy (96%).

Automatic inspection of solar panels based on Images stitching technique

Solar panels outer surface inspection is among the most challenging tasks which burden the solar plants supervisors. An automatic inspection based on image processing ensured by robots seems beneficial. In our approach, a robot equipped with an HD camera and a processing unit browses a set of solar panels (placed alongside). Afterward, the robot captures a series of successive images containing a set of targeted panels. Thereafter, the system generates a panoramic image of the panels which helps the said supervisors to check up the solar panels outer state i.e. (present of: dust, breaking, cracks,…) The proposed system is able to operate as an embedded system (ensured by Raspberry Pi card). It consists of two main stages: (1) images indexation and (2) Images stitching. As regards to the first stage, a series of solar panels images are captured successively and each of them is watermarked by a robust index, based on the convolutional encoding method and on the DCT blocks watermarking technique. Thereafter, the images undergo a brightness compensation as to obtain an accurate panoramic image. As for the second stage, the system starts by extracting the interest points (based on the SURF algorithm) of each captured image. Afterwards, the system implements a combination of Homography and RANSAC algorithm as to remove the mismatching points. And finally, each image undergoes a geometrical transformation according to the remaining matching points and subsequently the resulting images are stitched respectively as to obtain a panoramic image.

Solar panel monitoring using a video frames mosaicing

One of the most critical issues facing solar plants is to ensure an efficient monitoring of installations, such as localizing dust or broken panels. Filming with robots or drones is widely used, but it is difficult to exploit directly the captured videos. In order to facilitate the monitoring task, we propose a technique to prepare a large panoramic image view of the desired area from the captured video frames. In fact, an embedded system analyses the captured video sequence, and using a new approach of mosaicing technique based on SIFT, merges frames to prepare a panoramic image of a large solar plant area. The proposed technique facilitates the supervisor task to better locate where there are anomalies. The obtained experimental results, tested on a small solar installation, are promising.

A novel image detail preserving impulse noise removal algorithm

In this paper, we introduce a novel edge directed image detail preserving impulse noise removal algorithm. Our approach is based on local image directionality features. Local edge features are analysed on both a downsampled version of the image and the noisy image, and both soft an sharp edges are considered for selective noise removal. Experimental results on a set of standard images show our technique to be effective in removing salt-and-pepper noise even at high noise levels, to yield good image quality and to outperform a number of other noise removal techniques.

Sky status: A local analysis of ground based digital images

Cloud coverage obscures the sunlight, it can be identified as the major factor that significantly decreases the total irradiance received by photovoltaic panels, and then the power generation is intermittent. In this paper, and in order to predict the energy production by a solar installation, we analyze the acquired ground based images of the sky to extract its actual status (clear, partially cloudy, or totally cloudy). An embedded system is dedicated to simultaneously capture images and estimate the power generation. The accurate sky status is determined by comparing the brightness of the sun zone in the digital image to its surrounding area. The experimental results are promising and shows that the developed techniques can be easily embedded in solar plants to accurately predict the short time energy production.

Cloud tracking from whole-sky ground-based images

Tracking and predicting cloud motion was a key factor in getting the most out of solar PV installations. This paper presents a method of tracking cloud motion using ground-based omnidirectional images acquired by a catadioptric camera. This sensor offers wide field of vision that allows us a longtime of clouds tracking. Acquired images are processed to extract the relative position of the sun and tracking the cloud motions. The proposed tracking procedure is composed of three steps: segmentation, localization and detection of interest points of clouds then tracking cloud motions using an optical flow method, finally we determine the velocity vector. This technique can track displacement of each interest point in different speeds and directions to follow cloud path.