Abstract: A method for encoding a video stream includes obtaining images of a scene captured by a camera at a first resolution; identifying regions of interest (ROIs) in an image; adding, as part of an encoded video stream, a first video frame encoding at least part of the image at a second resolution lower than the first resolution; adding a second video frame marked as a no-display frame, and being an inter-frame referencing the first video frame with motion vectors for upscaling of the ROIs; adding a third video frame encoding the ROIs at a third resolution higher than the second resolution, and being an inter-frame referencing the second video frame.

Abstract: An image processing device generates an output image with masking of objects of classes. An input image is downscaled to an object detection image having a lower resolution than the input image and lower than the output image. The object detection image is input to an object detection module and confidence scores for pixel areas are received from the object detection module. Each confidence score indicates a probability that the pixel area relates to an object to be masked. Based on the input image, an intermediate image is generated having a higher resolution than the object detection image resolution and an adaptive masking threshold is set such that the greater the ratio between the output image resolution and the object detection image resolution, the lower the masking threshold. The output image is then generated by masking pixel areas of the intermediate image.

Abstract: A method of encoding images in a video, comprises: acquiring an original image from an image sensor of a video camera; encoding the original image using a generative image model, thereby obtaining a first encoded image; decoding the image to obtain a first decoded image; identify a region of interest (ROI) of the original image; for each ROI: perform an encoding quality check by comparing several reference points in the ROI of the original image against corresponding reference points in the ROI of the first decoded image, thereby obtaining a difference; if the difference is greater than a threshold, encoding the ROI using a non-generative image model, thereby obtaining a non-generative encoded image area; providing final encoded image data comprising a) the non-generative encoded image areas for the ROI having a difference greater than the threshold and b) the first encoded image for a remaining part of the original image.

Abstract: A method for processing of data by applying a first algorithm in a system comprising a server and a client device comprises at the client device: continuously receiving data; dividing the received data into M>1 subsets of data, each subset of data comprising, or being derived from, sensor data captured by a sensor during a capture time interval at a plurality of points in time; determining for each subset whether the processing of the subset should be performed by the client device or by the server, and upon determining that the processing should be performed by the client device, processing the subset into processed data by applying the first algorithm to the subset, and transmitting the processed data to the server, and upon determining that processing should be performed by the server, transmitting the subset of data to the server.

Abstract: A system for handling bandwidth shortage for transmission of encoded video frames. In a sending device, a long term reference frame which indicates bandwidth shortage is created. During a first time period, the reference frame is sent to the receiving device for storage. During a subsequent second time period, bandwidth shortage is determined in the sending device, wherein the bandwidth during the second time period is insufficient if the encoded video frames are encoded according to an encoding principle used at times without bandwidth shortage. During the second time period, inter encoded frames referencing the long term reference frame are sent, wherein each frame includes encoded blocks, and at least a subset of the blocks of the inter encoded frames are empty blocks such that the bit rate of the inter encoded frames is lower or equal to the bandwidth during the second time period.

Abstract: This disclosure relates to methods systems and non-transitory computer-readable storage mediums for distributing load in a multi-chip image processing unit for processing image data into processed image data. An example method comprises receiving first image data, analysing the first image data using a first algorithm, the first algorithm performing a set number of operations for a given size of image data input to the first algorithm, and outputs at least one characteristic of the first image data; using the at least one characteristic to estimate use of memory bandwidth in the first and second chip when processing the first image data into processed image data; and distributing processing of the first image data between the first and the second chip such that the estimated use of memory bandwidth is distributed evenly.

Abstract: A device and method merge a first candidate area relating to a candidate feature in a first image and a second candidate area relating to a candidate feature in a second image. The first and second images have an overlapping region, and at least a portion of the first and second candidate areas are located in the overlapping region. An image overlap size is determined indicating a size of the overlapping region of the first and second images, and a candidate area overlap ratio is determined indicating a ratio of overlap between the first and second candidate areas. A merging threshold is then determined based on the image overlap size, and, on condition that the candidate area overlap ratio is larger than the merging threshold, the first candidate area and the second candidate area are merged, thereby forming a merged candidate area.

Abstract: A method for masking of objects in a video stream, the method comprising: acquiring a video stream; detecting an object in the video stream; determining whether the detected object belongs to a foreground of the video stream indicative of moving objects or to a background of the video stream indicative of static objects; classifying the detected object to be of a specific type using a first classifier if the detected object is determined to belong to the foreground, and using a second classifier if the detected object is determined to belong to the background, the first classifier being different from the second classifier, and if the detected object is classified as being of the specific type of object, masking the object in the video stream.

Abstract: A device, and method of signing a video segment comprising one or more groups of pictures, GOPs, wherein each GOP comprises a header and one or more frames, are disclosed. For each of the one or more GOPs a GOP hash is produced and the GOP hash is digitally signed by means of a digital signature to produce a signed GOP hash. For each GOP except a last GOP of the one or more GOPs the respective signed GOP hash is saved in the header of a subsequent GOP. An additional GOP is added to the video segment after the last GOP of the one or more GOPs, wherein the additional GOP comprising a header and one or more frames. The signed GOP hash of the last GOP of the one or more GOPs is saved in the header of the additional GOP.

Abstract: Systems, methods and non-transitory computer-readable media transform raw image data into a video stream comprising a plurality of encoded image frames. To reduce memory bandwidth and power consumption, compression rates of sets of imaging data temporarily stored in memory during the transformation of raw image data into the video stream are adapted based on encoding configurations, on a pixel block level.

Abstract: A thermometric camera for determining a temperature of an object in motion comprises a bolometer sensor and circuitry. The bolometer sensor captures a sequence of image frames of the object while the object is moving. The circuitry executes: an object identifying function configured to identify an area corresponding to the object in each image frame of a series of image frames among the sequence of image frames; a combining function configured to combine the identified areas from each image frame in the series of image frames into a stacked image of the object, wherein pixel values in the stacked image of the object are estimated as a sum of pixel values of the corresponding pixels in the image frames in the series of image frames; and a temperature determining function configured to determine the temperature of the object in motion from pixel values in the stacked image of the object.

Abstract: A method of enabling a remote access to a first network device from a second network device includes the second device generating a data item signed by a network service trusted by the first device, and including a fingerprint of a public key of the second device. The signed data item is sent to the first device via a signalling service, SIGS, as part of a negotiation of terms for a peer-to-peer connection. The first device uses the received signed data item to verify that the terms it receives from the SIGS has not been tampered with by the SIGS, in order to prevent the SIGS from performing a Man-In-The-Middle attack. Various network devices and a network system are also provided.

Abstract: A method detects one or more occluded areas of a scene analysed by an object tracking system. The method includes building a map of one or more occluded areas in a scene. Building the map comprises running a re-identification algorithm on a video sequence to try to resume a lost object track. If the object track is successfully resumed, the method includes determining an area of the scene where the first object track is lost and an area of the scene where the first object track is resumed. A connection between the first and the second area of the scene is added the map such that the map identifies that an object track being lost in the first area of the scene has been resumed in the second area of the scene.

Abstract: A method for determining images plausible to have a false negative object detection comprises providing a group of historic trajectories, wherein each historic trajectory comprises a reference track that represents one or more historic tracks and comprises an object class of historic object detections that belong to the one or more historic tracks; performing tracking; performing object detection; for a determined track that does not match any determined object detection, comparing the determined track with reference tracks of historic trajectories for identifying a matching reference track; upon identifying a matching reference track, defining images of the determined track as being plausible to have a false negative object detection for the object class of the historic trajectory comprising the matching reference track; and upon not identifying a matching reference track, defining the determined track as a false positive track.

Abstract: A method for feature extraction of detected objects, comprising the steps of: receiving a plurality of images, each depicting an object detected by the object detecting application; concatenating the plurality of images into a composite image according to a grid pattern; feeding the composite image through a convolutional neural network (CNN) trained for feature extraction, wherein each convolutional layer of the CNN is configured to, while convolving input data to the convolutional layer using a convolutional filter: determine a currently convolved image of the plurality of images by determining a centre coordinate of a subset of the input data currently covered by the convolutional filter, and mapping the centre coordinate to the grid pattern; and selectively nullifying all weights of the convolutional filter that cover input data derived from any of the plurality of images not being the currently convolved image.

Abstract: A method for thermal image processing, comprises: acquiring a thermal image depicting a scene; obtaining a frequency distribution based on pixel intensities of the thermal image; processing the frequency distribution to determine whether the frequency distribution comprises a peak, caused by a hot object in the scene, which is separated from a thermal background of the scene by more than an intensity threshold; and in response to determining that the frequency distribution comprises the peak, processing the thermal image to suppress a ghost image of the hot object in the thermal image, wherein the ghost image is caused by internal reflections of radiation from the hot object in the thermal camera, and wherein the processing of the thermal image comprises: estimating a location of ghost image pixels forming the ghost image; and suppressing the ghost image in the thermal image by adjusting intensities of the ghost image pixels.

Abstract: A method, system and software for determine a color of a tracked object. Using a first video sequence and foreground objects detected therein, a color rendering metric may be determined for each area of a plurality of areas in the scene. Such color rendering metrics may then be used in case a tracked object is determined to have different colors in different images of a second video sequence, such that the colors detected in an area associated with a higher color rendering metric is selected over an area associated with a lower color rendering metric.

Abstract: The present system and method generally relate to the field of camera surveillance, and in particular to object re-identification in video streams captured by a camera.

Abstract: A method of processing digital video data comprises continuously capturing digital video data representing image frames. While capturing the digital video data, the digital video data is encoded into a sequence of encoded image frames, the sequence comprising key frames and delta frames, and storing the sequence of encoded image frames. It is then determined that the stored sequence of encoded image frames is to be entropy coded and, as a consequence, entropy coding the sequence of encoded image frames into an entropy coded sequence of image frames and storing the entropy coded sequence of image frames.