A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
FPGA SVM Safety system for excavator using SVM and Neuromachine based and FPGA
The present invention relates to an excavator safety system using an FPGA-based neuro machine and SVM. More specifically, images acquired by cameras are converted into one single surround-view image through a surround-view monitoring (SVM) algorithm to be displayed, and, as a result, accurate viewing angle information with a reduced blind spot to a driver is provided to the driver to prevent an accident as well as improve working efficiency and safety, and also, as objects and obstacles are omnidirectionally sensed from surroundings of a vehicle body and displayed to the driver, visual identification on whether an object is approaching can be possible, which can lead to a more effective reduction in accidents such as a collision and the like. Moreover, the SVM algorithm and a people-and-object detection (POD) algorithm are manufactured into FPGA modules to improve low power consumption and a processing speed while parallelly processing each image through a GPU of an AP to enable high-speed calculation, and also, an upper rotary body is mounted to be rotatable 360-degree from a lower driving body to visualize and provide rotation state information of an upper rotary body, when the surround-view image is generated, considering characteristics of an excavator in which forward, backward, left and right directions in accordance with a driving direction and forward, backward, left and right directions in accordance with a viewing direction are not matched, and, as a result, the information about the objects and the obstacles in the surroundings can be accurately and quickly provided to the driver.
본 발명은 SVM(Surround View Monitoring) 알고리즘을 이용하여 카메라들의 촬영에 의해 획득된 영상들을 하나의 서라운드-뷰 영상으로 변환하여 디스플레이 함으로써 운전자의 사각지대를 줄여 정확한 시야각 정보를 운전자에게 제공이 가능하여 안전사고를 미연에 방지할 수 있을 뿐만 아니라 작업효율 및 안전성이 개선되며, 차체 주변 전방위의 객체 및 장애물을 감지하여 운전자에게 표출시킴으로써 객체 접근 및 여부에 대한 시각적 식별이 가능하여 충돌 등의 안전사고를 더욱 효과적으로 절감시킬 수 있고, SVM 알고리즘 및 POD(People and Object Detection) 알고리즘이 FPGA 모듈로 제작됨에 따라 저전력 및 처리속도의 개선이 가능함과 동시에 AP의 GPU에서 각 영상을 병렬 처리하여 고속 연산이 이루어질 수 있으며, 상부 선회체가 하부 주행체로부터 360° 회전 가능하도록 탑재되어 주행방향에 따른 전후좌우 방향과, 시야방향에 따른 전후좌우 방향이 일치하지 않는 굴삭기의 특성을 감안하여, 서라운드-뷰 영상 생성 시, 상부 선회체의 회전상태 정보를 시각화하여 제공하도록 구성됨으로써 주변 객체 및 장애물에 대한 정보를 정확하고 신속하게 운전자에게 제공할 수 있는 FPGA 기반 뉴로머신과 SVM을 이용한 굴삭기 안전시스템에 관한 것이다.
FPGA SVM Safety system for excavator using SVM and Neuromachine based and FPGA
The present invention relates to an excavator safety system using an FPGA-based neuro machine and SVM. More specifically, images acquired by cameras are converted into one single surround-view image through a surround-view monitoring (SVM) algorithm to be displayed, and, as a result, accurate viewing angle information with a reduced blind spot to a driver is provided to the driver to prevent an accident as well as improve working efficiency and safety, and also, as objects and obstacles are omnidirectionally sensed from surroundings of a vehicle body and displayed to the driver, visual identification on whether an object is approaching can be possible, which can lead to a more effective reduction in accidents such as a collision and the like. Moreover, the SVM algorithm and a people-and-object detection (POD) algorithm are manufactured into FPGA modules to improve low power consumption and a processing speed while parallelly processing each image through a GPU of an AP to enable high-speed calculation, and also, an upper rotary body is mounted to be rotatable 360-degree from a lower driving body to visualize and provide rotation state information of an upper rotary body, when the surround-view image is generated, considering characteristics of an excavator in which forward, backward, left and right directions in accordance with a driving direction and forward, backward, left and right directions in accordance with a viewing direction are not matched, and, as a result, the information about the objects and the obstacles in the surroundings can be accurately and quickly provided to the driver.
본 발명은 SVM(Surround View Monitoring) 알고리즘을 이용하여 카메라들의 촬영에 의해 획득된 영상들을 하나의 서라운드-뷰 영상으로 변환하여 디스플레이 함으로써 운전자의 사각지대를 줄여 정확한 시야각 정보를 운전자에게 제공이 가능하여 안전사고를 미연에 방지할 수 있을 뿐만 아니라 작업효율 및 안전성이 개선되며, 차체 주변 전방위의 객체 및 장애물을 감지하여 운전자에게 표출시킴으로써 객체 접근 및 여부에 대한 시각적 식별이 가능하여 충돌 등의 안전사고를 더욱 효과적으로 절감시킬 수 있고, SVM 알고리즘 및 POD(People and Object Detection) 알고리즘이 FPGA 모듈로 제작됨에 따라 저전력 및 처리속도의 개선이 가능함과 동시에 AP의 GPU에서 각 영상을 병렬 처리하여 고속 연산이 이루어질 수 있으며, 상부 선회체가 하부 주행체로부터 360° 회전 가능하도록 탑재되어 주행방향에 따른 전후좌우 방향과, 시야방향에 따른 전후좌우 방향이 일치하지 않는 굴삭기의 특성을 감안하여, 서라운드-뷰 영상 생성 시, 상부 선회체의 회전상태 정보를 시각화하여 제공하도록 구성됨으로써 주변 객체 및 장애물에 대한 정보를 정확하고 신속하게 운전자에게 제공할 수 있는 FPGA 기반 뉴로머신과 SVM을 이용한 굴삭기 안전시스템에 관한 것이다.
FPGA SVM Safety system for excavator using SVM and Neuromachine based and FPGA
FPGA 기반 뉴로머신과 SVM을 이용한 굴삭기 안전시스템
CHO SEUNG PYO (author) / LEE HYE SEONG (author)
2021-12-14
Patent
Electronic Resource
Korean
IPC:
E02F
Baggern
,
DREDGING
/
B60K
Anordnung oder Einbau von Antriebseinheiten oder von Kraft- bzw. Drehmomentübertragungen in Fahrzeugen
,
ARRANGEMENT OR MOUNTING OF PROPULSION UNITS OR OF TRANSMISSIONS IN VEHICLES
/
B60R
Fahrzeuge, Fahrzeugausstattung oder Fahrzeugteile, soweit nicht anderweitig vorgesehen
,
VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
/
B60W
CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION
,
Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion
FPGA SVM Safety system for excavator using SVM and Neuromachine based and FPGA
European Patent Office | 2022
FPGA-based interface model and inter-FPGA communication
| British Library Online Contents | 2010
Invariant-Based Safety Assessment of FPGA Projects: Conception and Technique
| BASE | 2021