Focusing on FPGA image solution customization

brief introduction:
This product is based onFPGAUsing the high-speed parallel characteristics of programmable logic devices, as well as its rich internal multipliers andRAMThe resource has completed the collection of venous images and the hardening of convolutional neural network algorithms, ultimately adoptingTIofDLPCompleted the design of a handheld venous projector and solved medical difficulties.
Text Introduction:
Intravenous sampling or drug injection is an essential means of treatment for doctors and patients. But for young and weak babies, the veins are very thin and difficult to find; Obese adults with thick sebum hinder the localization of veins, which not only troubles doctors and nurses, but also easily causes secondary harm to patients.
To address such medical challenges, we are based onFPGADesigned a handheld venous projector:VN-DP001. This device usesCylcone IV FPGAAs the main control device, under the projection of near-infrared lightCMOSSensors capture images; Continuing to utilizeFPGAWith the high-speed parallel feature, we use a large window multi-scale convolutional neural network to calculate the light intensity data partially absorbed by hemoglobin. Finally, through a series of image enhancement algorithms, weTIofDLPRe project onto the arm and overlay onto the veins that are not easily visible to the naked eye. The overall flowchart is as follows:

Product selectionCyclone IV FPGAbyEP4CE40F23C8NThe chip has40K LEs,232individual9bitMultiplier, and1.1KofRAMThe final resource utilization rate is shown in the figure. Thanks to thisFPGAWe use dedicated multipliers and abundant wiring resources40Frequency doubling, achieving equivalence9280The reuse of multiple multipliers ultimately achieves the high performance and cost-effectiveness of this product.
This product solves medical problems and brings good news to patients. As of now, this product has been cumulatively released both domestically and internationally4In the year, it sold in both domestic and international medical markets and continued to be favored by hospitals and related institutions.
along withFPGAWith the continuous improvement of computing power and the increasing demand for medical services, in the foreseeable future, handheld vein projectors with high resolution, high brush speed, and even precise depth detection will be the direction of development for such products. We will also continue to work hard and embraceintel FPGAContinue to deepen our cultivation in the medical field, committed to solving more problems and realizing the value of technology.
Function
Programme
SZOVS,Born for FPGA images…
Buy NowDetailed introduction

简介:
本产品基于FPGA,采用可编程逻辑器件的高速并行特性,以及其内部丰富的乘法器与RAM资源,完成了静脉图像的采集,以及卷积神经网络算法的硬化,最终采用TI的DLP,完成了手持静脉投影仪的设计,解决了医疗难题。
文字介绍:
静脉采样,或药物注射,是医患治疗中的必备手段。但年幼体弱的小宝宝,静脉很细很难找到;体型肥胖的成年人,皮脂肥厚阻碍了静脉的定定位,这不仅困扰了医生护士,也容易给患者带来了二次伤害。
为了解决此类医疗难题,我们基于FPGA,设计了一套手持静脉投影仪:VN-DP001。该设备采用Cylcone IV FPGA作为主控器件,在近红外光的投射下,通过CMOS传感器采集图像;接着利用FPGA高速并行特性,我们采用大窗口多尺度卷积神经网络,计算部分被血红蛋白吸收后光强数据,最后通过一系列图像增强等算法后,由TI的DLP重新投射到手臂,叠加到肉眼不易见的静脉上。整体的流程框图,如下所示:

产品选用的Cyclone IV FPGA为EP4CE40F23C8N,该芯片具有40K LEs,232个9bit的乘法器,以及1.1K的RAM资源,最终的资源利用率如有图所示。得益于该FPGA专用的乘法器以及丰富布线资源,我们采用40倍频,实现了等效9280个乘法器的复用,最终实现了本产品高性能,高性价比的特性。
本产品解决了医疗的难题,也带来了患者的福音。截至目前,本产品已经在国内外累积发行4年,销售国内外医疗市场,并且持续的受到医院以及相关机构的亲睐。
随着FPGA计算力的不断提高,以及日益递增的医疗需求,在可见的未来,高分高刷,甚至可以精确探测深度的手持静脉投影仪,将是此类产品发展的方向。我们也将持续努力,拥抱intel FPGA,在医疗领域继续深耕,致力于解决更多的难题,兑现技术的价值。

302, Building 17, Nanshan Ruiyuan, Xili Longzhu 3rd Road, Nanshan District, Shenzhen City, Guangdong Province
19926595670
cb@szovs.com
WeChat
Skype
Official WeChat
website
FPGA website
Official account
Official WeChat public account
Official account
Official account
Skype
Skype
Mailbox
cb@szovs.com
Comeback