Narrow down by specifying conditions
現在1887件がヒットしています。check
When considering an image processing system using Jetson, many people may have the following concerns.
[Two concerns frequently asked by customers]
1. I don't know what kind of development tools should be used to create and implement a program that can run on Jetson
2. I don't know the specifications of Ubuntu, software installation is troublesome, and I want to start verification immediately, but it's hard to get started
Therefore, this time, we would like to introduce a kitting set that combines HALCON, an image processing library, a Baslar camera, and an edge terminal equipped with NVIDIA Jetson™.
This set will be delivered with the OS and software already installed. Therefore, troublesome installation work is not required, and deep learning operation verification and image processing development using HALCON can be performed immediately. In addition, the required performance and specifications differ depending on the application, so we have prepared two types of kits.
Let me introduce you briefly.
What is HALCON?
The image processing library HALCON is a pioneering tool that allows you to build an efficient image processing system by combining 2,000 library operators. The created script type program can be automatically converted to Visual C++, Visual Basic or Visual C# code, and can be easily incorporated into the customer's image processing system. HALCON is currently used in tens of thousands of systems in Japan alone, demonstrating the high reliability of the HALCON library.
Source: Links Co., Ltd. https://linx.jp/product/mvtec/halcon/
Application example of HALCON
Case 1: Acceptance/rejection judgment of blood collection tube caps
Product quality and traceability have been improved by using deep learning to classify defects that are difficult to quantify, such as breakage and peeling of barcode stickers, and by combining this with code recognition functions.
Case 2: License plate recognition
Recognizing objects outdoors is extremely difficult because the appearance of objects changes depending on the weather and time of day, but we have combined rule-based image processing and deep learning to build flexible logic.
What are the advantages of combining HALCON and Jetson?
The combination of HALCON and Jetson is the optimal solution that enables edge AI to be implemented flexibly.
In order to realize image edge AI, it is essential to have an image processing library, affinity with your own system, and edge computing that can perform high-speed calculations.
The kit introduced this time achieves flexible image edge AI that meets customer needs by combining HALCON, an image processing library that can automatically convert code, and NVIDIA Jetson, which specializes in parallel computing technology. can. In addition, image processing algorithms built on Windows, Linux, etc. can be easily deployed on Jetson and their operation can be evaluated.
2 types of validated sets are available
HALCON, an image processing library used in tens of thousands of systems in Japan, and a Basler camera are kitted to an edge terminal equipped with NVIDIA Jetson.
There is no need for the customer to install any OS or software. After purchase, you can immediately start operation verification of image processing deep learning using HALCON and operation verification of image processing development application.
1. EDGEMATRIX AI-BOX with NVIDIA Jetson™ TX2 Module
This set uses the EDGEMATRIX AI-BOX.
AI-BOX with NVIDIA Jetson TX2 Module equipped with NVIDIA Pascal™ architecture-based GPU. Equipped with TX2, which is ideal for intelligent edge devices, it enables wider and more complex neural networks.
Source: https://edgematrix.com/
set accessories
|
set content |
accessories |
Model |
| Basler ace series GigE camera | GigE camera body | acA1440-73gc |
| GigE cable 5m | GEV-C6-L05-ACE | |
| lens | C125-0418-5M | |
| 12V power supply unit (with trigger cable 3m) |
PS12V-L03-Trig-ACE | |
| AC adapter cable | ST-Power Cable | |
| tripod mount adapter | ACE-TPM | |
| EDGEMATRIX AI-BOX | AI-BOX body | EHC-J2-301 |
| Image processing library HALCON | HALCON20.11 Development Package + Deep Learning | HALCON20.11-DEEP |
Product spec
Basler camera spec table
|
item |
acA1440-73gc Specifications |
|
Number of pixels (pix) |
1440x1080 |
|
number of pixels |
1.6MP |
|
sensor type |
Progressive Scan CMOS |
|
sensor name |
Sony IMX273 |
|
Image element size (inch) |
1/2.9" CMOS |
|
shutter |
global |
|
Pixel size (um) |
3.45um (H) x 3.45um (W) |
|
フレームレート(fps) |
73 |
|
monochrome/color |
Color |
|
video output type |
GigE |
|
video output format |
10, 12 bit |
|
sync mode |
External trigger, software trigger or free run mode |
|
exposure control |
Programmable mode, external trigger control |
|
External size (L x W x H) |
29mm x 29mm x 29mm |
|
Power-supply voltage |
PoE or 12-24VDC |
|
lens mount |
C-mount |
Basler lens specification table
|
item |
C125-0418-5M Specifications |
|
fixed focal length |
4mm |
|
Aperture range |
F1.8 to F22 |
|
resolution |
5 megapixel |
|
lens mount |
C-mount |
EDGEMATRIX AI-BOX spec table
|
Specifications |
Edge AI Box - Indoor : EHC-J2-301 |
|
Construction |
Aluminum die-casting and steel, IP40 |
|
Core Module |
NVIDIA Jetson TX2 |
|
GPUs |
NVIDIA Pascal, 256 CUDA cores |
|
CPU |
ARM Cortex-A57@2GHz + NVIDIA Denver2@2GHz |
|
Memory |
8GB 128-bit LPDDR4@1866MHz onboard |
|
System I/O Outlets |
1x USB 2.0 |
|
PoE |
- |
|
Storage |
1 x M.2 w/ PCIe 3.0 (support M.2 NVMe SSD) : STD128GB |
|
LTE |
4G (LTE) / 3G (800/900/1800/2100MHz) |
|
Wifi |
802.11a/b/g/n/ac |
|
Power supply |
12VDC screw type connector |
|
Operating temperature |
-10℃ to +50℃ |
|
Humidity |
10% ~ 95%, non-condensing |
|
Vibration Endurance |
1 Grms w/ M.2 (5 ~ 500Hz, X, Y, Z directions) |
|
Dimensions |
141.6mm (5.57") (W) x 106mm (4.17") (D) x 55mm (2.17") (H) |
|
Weight (net/gross) |
0.8kg (1.76lb)/1.5Kg (3.31lb) |
|
Certificate |
CE & FCC Compliance |
2. AVerMedia BOX PC with NVIDIA Jetson AGX Xavier Module
BOX PC with NVIDIA Jetson™ AGX Xavier™ developed as a powerful AI computer for autonomous machines. It delivers performance comparable to GPU workstations and is intended for AI machines that require maximum computation.
Jetson AGX Xavier delivers 20x more compute performance and 10x more power efficiency than Jetson TX2.
Source: https://www.avermedia.co.jp/jp
set accessories
|
set content |
accessories |
Model |
|
Basler ace series |
USB camera body |
acA1440-220uc |
|
USB cable |
USB3_Cable_5M_S |
|
|
lens |
C125-0418-5M |
|
|
tripod mount adapter |
ACE-TPM |
|
|
AVerMedia BOX PC |
BOX PC body |
EA713-AAMN-3PC0 |
|
Image processing library HALCON |
HALCON20.11 Development Package + Deep Learning |
HALCON20.11-DEEP |
Product spec
Basler camera spec table
|
item |
acA1440-220uc Specifications |
|
Number of pixels (pix) |
1440x1080 |
|
number of pixels |
1.6MP |
|
sensor type |
Progressive Scan CMOS |
|
sensor name |
Sony IMX273 |
|
Image element size (inch) |
1/2.9" CMOS |
|
shutter |
global |
|
Pixel size (um) |
3.45um (H) x 3.45um (W) |
|
フレームレート(fps) |
220 |
|
monochrome/color |
Color |
|
video output type |
USB3.0 (USB3 Vision) |
|
video output format |
8, 12 bit |
|
sync mode |
External trigger, software trigger or free run mode |
|
exposure control |
Programmable mode, external trigger control |
|
External size (L x W x H) |
29mm x 29mm x 29mm |
|
Power-supply voltage |
USB 3.0 interface |
Basler lens specification table
|
item |
C125-0418-5M Specifications |
|
fixed focal length |
4mm |
|
Aperture range |
F1.8 to F22 |
|
resolution |
5 megapixel |
|
lens mount |
C-mount |
AVerMedia BOX PC spec table
|
Type |
Box PC with AVerCooler WavieFin passive cooling systemEA713-AAMN-3PC0 |
|
NVIDIA GPU SoC Module Compatibility |
NVIDIA® Jetson AGX Xavier™ 32GB |
|
Networking |
2x GbE (RJ-45) |
|
Display Output |
1x HDMI type A, maximum resolution: 3840x2160 at 60Hz |
|
temperature |
Operating Temperature -20°C~65°C (MODE_30W_ALL, No-Air-Flow) |
|
USB |
1x USB 2.0 Micro-B for BSP installation only |
|
Storage |
32GB eMMC 5.1 |
|
RS-485 |
1x RS-485 |
|
CAN bus |
1x CAN bus with transceiver |
|
GPIO Expansion |
40 pins: 1x UART, 2x I2C, 1x CAN (W/O transceiver), 5x GPIO |
|
User Expansion |
1x M.2 Key E2230 for Wi-Fi module (PCIe Gen4x 1) |
|
Input Power |
12V/5A |
|
buttons |
Power and Recovery (Each button has a RGB tri-color LED) |
|
RTC Battery |
Support RTC battery and Battery Life Monitoring by MCU |
|
Chassis Dimension/Weight |
W:196mm x L:232mm x H:60mm |
|
Certifications |
CE, FCC |
price
Please contact us for the price of this product.
Contact Us
If you would like a detailed explanation of the kitting set featured in this article, or if you have any concerns about selection, please contact us here.
![[Lecture video now available] Thumbnail image of the event report "I want to know now! Frontline seminar on the use of industrial drones and UGVs"](/business/semiconductor/articles/013_133786_thumb_1.png)
