Daegu Gyeongbuk Institute of Science and Technology

Korea Institute of Industrial Technology

Soon O Kwon

Jin Hee Lee

Eui Chun Lee

Woong Jae Won

Jong Hun Lee

Byung In Moon

Ho Young Lee

Kyeong Ryeol Bae

Kang Won Lee

Tae Hun Kim

Kumar Ajay

Su Woong Lee

Jae Suck Kim

Jae Hyeong Park

The present invention relates to a technical idea for providing a delivery service on regular and irregular roads using autonomous vehicles. More specifically, the present invention relates to technology in which, on a regular road, a lead vehicle and at least one droid vehicle are coupled to each other and a delivery service is provided based on autonomous driving; and on an irregular road, the coupling between the lead vehicle and the droid vehicle is automatically released and the droid vehicle provides a delivery service by remotely controlling the driving of the droid vehicle by the lead vehicle in the last mile delivery section corresponding to the irregular road. According to one embodiment of the present invention, a system for providing a delivery service using autonomous vehicles may provide a delivery service on an irregular road where entry of normal vehicles is not allowed and a regular road where entry of small and low-speed vehicles is not allowed and may include a droid vehicle for providing a delivery service using limited autonomous driving performance in a last mile delivery section corresponding to the irregular road; and a lead vehicle for providing a delivery service based on autonomous driving on the regular road, transporting the droid vehicle by being coupled to the droid vehicle on the regular road, and remotely controlling driving of the droid vehicle after being separated from the droid vehicle in the last mile delivery section.

Delivery service system and method using autonomous vehicles

An apparatus and method for data fusion between heterogeneous sensors are disclosed. The method for data fusion between the heterogeneous sensors may include identifying image data and point cloud data for a search area by each of a camera sensor and a LiDAR sensor that are calibrated using a marker board having a hole; recognizing a translation vector determined through calibrating of the camera sensor and the LiDAR sensor; and projecting the point cloud data of the LiDAR sensor onto the image data of the camera sensor using the recognized translation vector to fuse the identified image data and point cloud data.

Device and method for data fusion between heterogeneous sensors

Disclosed are a real-time object detection method and an apparatus therefor. An object detection apparatus may receive an input image, may extract a first feature map from the input image, may detect an object included in the input image based on the first feature map, may extract, from the input image, a second feature map having resolution higher than the resolution of the first feature map, may extract a third feature map from the second feature map based on the region of the detected object, and may redetect the object based on the first feature map and the third feature map.

Method and device for detecting object in real time by means of deep learning network model

A method for manufacturing a distributed straight-angle armature winding according to the present invention comprises a step (a) of preparing a raw material in which at least a pair of intra-slot winding areas and an end winding area provided between the pair of intra-slot winding areas are continued, a step (b) of forming an inclined surface by processing a contact surface in contact with an inner surface of a slot of at least a stator of the circumferences of the pair of intra-slot winding areas, and a step (c) of bending a space between the end winding area and the intra-slot winding area.

Distributed straight-angle armature winding, motor comprising same, and method for manufacturing same

A flat-angled coil having a three-dimensional shape for maximizing a space factor according to the present invention comprises a unit coil comprising multiple layers which are provided to surround the peripheries of teeth of a core provided on an electric motor and laminated along the longitudinal direction of the teeth of the core so as to be integrally connected to each other continuously. The unit coil is formed such that at least a part of an adjacent portion thereof, which is adjacent to another unit coil provided in the same slot formed between the surrounded teeth of the core and the teeth of an adjacent core, is non-parallel with the side surfaces of the surrounded teeth of the core.

Flat-angled coil having three-dimensional shape for maximizing space factor and electric motor comprising same

Disclosed is a method for manufacturing an MSO coil, comprising: a pressing step of forming a bent surface on a part of a unit coil layer, which has a ring shape such that both ends thereof face each other, thereby endowing both ends of the unit coil layer with a height difference; a fixing step of connecting and fixing a plurality of unit coil layers to each other, each unit coil layer having the bent surface formed thereon, such that the first end of both ends of a unit coil layer having the bent surface formed thereon contacts the second end of both ends of another unit coil layer having the bent surface formed thereon; and a bonding step of bonding connection parts defined by contact of the first and second ends of each of the plurality of unit coil layers that are connected and fixed to each other.

Method for manufacturing MSO coil and device for manufacturing same

A method and apparatus for performing parallel box filtering in region based image processing is provided. The method of performing parallel box filtering in region based image processing includes selecting parallel pixel blocks having a size of M&#xd7;N to be operated in parallel, selecting pixels included in a mask region having a size of K&#xd7;L that is a parallel pixel operation region and a first region having a size of (M+K&#x2212;1)&#xd7;(N+L&#x2212;1) that corresponds to overlapping masks of M*N, selecting pixels of a second region commonly included in the mask regions of M*N and storing the results of operation with respect to the pixels of the second region as repeated operation values, and operating block operation values with respect to the parallel pixel blocks using the repeated operation values.

Growing community of inventors

Daegu, South Korea

Soon O Kwon