Abstract: A method prunes a prediction-coded video sequence and a corresponding pruning device. The pruning method comprises: obtaining one or more global motion values for the video sequence, wherein each global motion value represents an amount of motion in a scene depicted by a set of successive image frames in the video sequence; for a set of successive image frames having a global motion value being below a global motion threshold, identifying an immediately following group of pictures, GOP; obtaining a skip frame having the same frame configuration as the image frames of the identified GOP; replacing the initial intra frame of the identified GOP with the obtained skip frame; and configuring the skip frame and the remaining image frames of the identified GOP to form a continuation of a frame numbering of an immediately preceding GOP.

Abstract: A computer-implemented method for segmentation and object detection in images captured by a camera having zoom capability comprising: obtaining images from the camera and obtaining a zoom factor, the zoom factor representing a current zoom level applied by the camera when capturing the images; selecting, based on the zoom factor, one neural network from a plurality of neural networks configured to segment the images and detect objects in the images. The plurality of neural networks operate at different image resolutions, wherein a higher zoom factor corresponds to selecting a neural network operating at a lower image resolution, and a lower zoom factor corresponds to selecting a neural network operating at a higher image resolution; down-scaling the images to the image resolution required by the selected neural network; applying the selected neural network to segment the down-scaled images and detect the objects in the down-scaled images.

Abstract: Methods, software, devices and systems for video scrubbing enable a client device to retrieve images for scrubbing based on a user-requested time along a video timeline of a video stored in a server. The client device checks if a cached image meets specified conditions, including a timestamp within a precision margin around the requested time. The precision margin scales with the timeline length, providing a smaller margin for shorter timelines and a larger margin for longer timelines. If a relevant cached image is found, it is retrieved; if not, an image with a highest relevance score within the precision margin is fetched from the server and stored in memory.

Abstract: A method for controlling an image stabilization (IS) system of an image capturing device. The IS system comprises an optical image stabilization (OIS) system that compensates for vibrational movement based on a motion signal from the image-capturing device's motion sensor. The OIS system has a first control parameter with an adjustable value to adapt responsiveness to vibrational movement, the value being adjustable between a lower responsiveness and a higher responsiveness of the OIS system. While the OIS system is in an active state and performs OIS, it determines a signal-to-noise estimate for the motion signal, and also determines the value of the first control parameter based on the signal-to-noise estimate, comprising setting the first control parameter to the first value responsive to determining that the signal-to-noise estimate is in a lower range, and to the second value responsive to determining that the signal-to-noise estimate is in an upper range.

Abstract: A method for streaming live coded video to one or more clients comprises: detecting a plurality of video stream requests; obtaining maximum bitrate values that return video streams are allowed to have; calculating requested quality of experience, QoE, values; determining a number of return video streams being lower than the plurality of video stream requests, and having bitrates such that each of the video stream requests can be matched with a return video stream, wherein the bitrates of the return video streams are determined such that a cumulative difference between the QoE value of each video stream request and a calculated return QoE of a matching return video stream is minimized; obtaining the return video streams; and streaming the return video streams. A streaming controller for controlling streaming of live coded video is also disclosed.

Abstract: A computer-implemented method to adjust a process noise of an object tracker in a camera is configured to monitor a scene including moving objects, the object tracker uses a motion model associated with process noise, the method comprising: determining, using the object tracker, object tracks of moving objects travelling along a path in the scene for a time period using a level of process noise, evaluating at least one spatial parameter of the determined object tracks in relation to a predetermined criterion, repeating the steps of determining object tracks and evaluating at least one spatial parameter for object tracks during subsequent time periods using increased process noise for each iteration until the predetermined criterion is fulfilled or a maximum process noise limit is reached, and storing data indicating the final process noise.

Abstract: There is provided an electronic access control system for a device. An OACS gateway is configured to obtain, via an OACS client, a request for the OACS gateway to issue a service ticket for a user agent. The OACS gateway is configured to retrieve, via an OIDC client, a profile for the user agent and an access control list for a profile and a device from the OIDC provider. The OACS gateway is configured to, responsive to having confirmed, by checking an access control list for the profile and the device, and by checking the profile itself, that the profile fulfils a permission to access a service on the device, generate a service ticket that validates the user agent for a one-time access to the service. The OACS gateway is configured to provide, via the OACS client, the service ticket towards the user agent.

Abstract: A method for intra refresh encoding of a video stream includes encoding the video stream using a first intra refresh pattern; determining that a time-variation of bits-per-frame for encoding the video stream using a second intra refresh pattern different from the first intra refresh pattern is lower than that for encoding the video stream using the first intra refresh pattern, and switching to encoding the video stream using the second intra refresh pattern. A corresponding device configured to perform the method is also provided, as well as a corresponding computer program and computer program product.

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: Anchor-based object detection system and method that identify potential false positive detections among three detection boxes in a same image frame. Each detection box has a predicted IoU score representing confidence that it captures an object. The system determines overlap of a second box with a first and third box, where the second is positioned between them. It identifies the second box as a potential false positive if it determines the IoU score of the second box is below a set threshold and if corresponding reference points within each box is determined to be substantially aligned, with the second box's reference point close to an alignment line defined by the first and third boxes.

Abstract: A camera device comprising a first printed circuit board (PCB) segment provided with a first grounding contact surface, a second PCB segment provided with an opening defined by a portion of the second PCB segment forming a second grounding contact surface, a chassis made of an electrically conductive material, a clip made of an electrically conductive material and snap fitted to the chassis; wherein the chassis is configured to support an image sensor and the first and second PCB segment, wherein the first PCB segment is arranged between the chassis and the second PCB segment, wherein the clip comprises a first contact portion abutting the first grounding contact surface of the first PCB segment; and a second contact portion extending through the opening of the second PCB segment and engaging the second grounding contact surface thereof.

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 screw joint arrangement that allows a second member to be fixed or movable relative to a first member. The first member includes at least one screw-receiving opening. The second member includes a bore having, in order from the first member side, a first threaded section, a non-threaded section, and a second threaded section. A screw with a head and a shaft carrying a threaded portion cooperates with these features to provide two selectable states. In the fixed state, a screw-receiving opening is aligned with the bore and the screw's threaded portion engages the second threaded section, clamping the first member between the screw head and the second member. In the movable state, rotation advances the screw until its threaded portion lies within the non-threaded section while the shaft passes through the first threaded section, thereby releasing the clamp and retaining engagement between the members for adjustment.

Abstract: A method for indicating clock drift of an internal clock comprised in a sensor device comprises: capturing, by the sensor device, first data frame; associating, by the sensor device, the first data frame with a first point in time indicating a point in time of the internal clock when the first data frame being captured; receiving, by the sensor device via a network protocol for clock synchronization, time synchronisation data; determining, by the sensor device using the time synchronisation data, a reference point in time indicating a point in time of a reference clock when the first data frame being captured; determining, by the sensor device, an offset between the first point in time and the reference point in time; and upon the offset exceeding a threshold offset, associating, by the sensor device, the first data frame with data indicating the offset.

Abstract: A method for determining occupied positions in a space starts by receiving detections of object movement from a radar monitoring the space. Each detection is associated with a position in the space and a time point when the detection was made. The method accumulates detections that have a time point within a first time period of predefined duration, and identifies a set of clusters of detections in the space by analyzing similarity in position of the accumulated detections. Any cluster whose detections are temporally distributed within a proportion of the first time period which is below a predefined proportion threshold are then removed from the set of clusters. The positions in the space that correspond to the set of clusters are determined to be occupied during the first time period.

Abstract: Techniques update background layers during encoding of scene, performed by an image processing device. The scene is encoded based on classifying objects depicted in the scene as either foreground or background. The background is divided into ordered background layers where each background layer is associated with a respective depth model. The method comprises detecting) a change in an image portion of one background layer. The method comprises calculating a difference between the image portion and a corresponding image portion of a background layer ordered behind the one background layer. The method comprises selecting, when the difference is smaller than a threshold, the background layer ordered behind the one background layer to represent the image portion.

Abstract: An image-processing device generates a three-dimensional model of a background scene of the image frame based on three-dimensional information about the background scene. The image-processing device defines a three-dimensional bounding box of the object in the three-dimensional model. The image-processing device defines a centre coordinate in the three-dimensional model and a colour value of surface elements of the three-dimensional bounding box. The image-processing device determines a three-dimensional coordinate of a surface in the three-dimensional model which reflects light from a surface element into the camera, by tracing rays from the centre coordinate and based on a normal of the surface. The image-processing device further identifies a first pixel in the image frame corresponding to the three-dimensional coordinate and detects the reflection of the object.

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.

Abstract: A device and a method for buffering a graphical overlay to be applied to an image is disclosed. A graphical overlay description specifying content, size, and position in the image of a graphical element of the graphical overlay is obtained, and the graphical overlay is divided into a plurality of sequential line fragments. For each line fragment it is determined, using the graphical overlay description, whether the line fragment overlaps a part of the graphical element. On condition that the line fragment overlaps a part of the graphical element, information representing the part of the graphical element is buffered in a buffer memory for the line fragment. On condition that the line fragment does not overlap any part of the graphical element, a run-length coding representing identical pixels is buffered in the buffer memory for the line fragment.