Early fire forest detection is crucial for fast and effective intervention. Many research have been done on this subject starting by sensor based systems and arriving to image processing which leverage the computer vision advancements. Our work refers to one of the latest algorithms in forest fire detection: YOLO. We present in this paper a detailed description of the architecture of the YOLO algorithm with an emphasis to the YOLOv6 which is the latest version of the YOLO algorithms. The performance of the studied algorithm is evaluated on a personal database containing 28334 images, with 10534 forest fire images and 17800 non-fire images. The experimental results of applying the YOLOv6 proved the efficiency of the method in fast and accurate fires detection even in large size images and low resolutions. This result makes the studied algorithm so suitable for both satellite and ground based images analysis.

With the start of 2020, the world witnessed the spread of Coronavirus disease (COVID-19). We aim in this work to employ artificial intelligence (AI) to develop a computer-aided diagnosis system (CAD) in order to automatically detect COVID-19 cases and differentiate them from normal and community-acquired pneumonia (CAP) cases through the use of lung Computed Tomography (CT) images and then evaluate its performance. Deep residual learning offers a wide variety of algorithms that helps in classification problems. We apply in this work a ResNet50 based model to recognize Covid-19 cases. Extensive analysis based on an international dataset (24256 images of 304 patients) proved that the ResNet50-optimized model can recognize COVID-19 through the use of CT images with 82% accuracy, 90% recall, 65% precision, and 76% of F1.Score.

Fires have become a more serious hazard to people's lives, property, and environment. Compared with the traditional techniques of fire detection, image technologies play a very promising role to overcome the problem of high false alarm rate. However, a major issue with these methods is their fastidious and long-time generation. In fact, the implemented algorithms are often produced using multi-feature technique, including chromatic characteristics, dynamic features, texture features and contour features. Therefore, we provide, in this paper, a study of some supervised machine learning algorithm for fire and smoke images recognition, and we compare it to a proposed model based on convolution neural network (CNN) algorithm. To do this, we consider a proper database composed by a total of 28334 images classified into three categories: 7329 fire images, 9205 smoke images and 11800 other images.

Detecting lymphoma lesions in the whole body is tedious and so much time consuming. In this paper, we propose a semi-automatic lymphoma lesion detection based on Chan-Vese algorithm to help doctors in their diagnosis. In addition, this study will helps doctors in the ADC measurement of the lymph nodes. The proposed algorithm is applied on real DW-MRI images obtained from 1.5T MR scan. To evaluate the proposed method, we compared the obtained results with those obtained with the region growing algorithm (RG). This evaluation is basically based on: sensitivity, specificity, accuracy, precision, and recall. The obtained values for these parameters confirm the efficiency of the proposed method especially for the accuracy which reaches 99.94% with query times.

Correction for internal gamma-ray attenuation is an important aspect in single photon emission computed tomography (SPECT). It is generally realized through the use of an attenuation map derived from X-ray CT data. In this paper we present a new method of creating an attenuation map derived from X-ray CT data where we extract the lungs by segmentation and we multiply it by the Hounsfield units (HU). This method is applied on a database of 10 thoracic SPECT exams. The evaluation of the obtained results shows that the proposed method increases the accuracy, the structural similarity index and the peak of signal to noise ratio of the obtained image comparing to the methods of Shabnam, Nasr and Saha. In addition, it guarantees a lower value of the mean squared error and lower time in execution.

Sometimes, computed tomography (CT) examinations need to be repeated. This may generate adverse effects on patients. To avoid it, an efficient reconstruction technique should be applied. This paper presents a qualitative and quantitative comparative study of four iterative algorithms (Algebraic Reconstruction Technique (ART), Maximum Likelihood Expectation Maximization (MLEM), Ordered-subsets expectation maximization (OSEM) and Simultaneous algebraic reconstruction technique (SART)). The four techniques are applied on a ‘dicom’ lung computed tomography image. Qualitatively, we can not differentiate between the reconstructed images. They are almost the same for all the methods. But, qualitatively, the best performance was observed with OSEM algorithm which provides the best quality image according to practically all the computed evaluation criteria. Moreover, OSEM insures this best performance in shorter processing time ranging from two to three times less compared to the other methods.