Lei Jia

ACF-COG interconnection conductivity inspection system using conductive area

The liquid crystal display (LCD) module package mainly adopts the chip-on-glass (COG) process, in which the driver IC is assembled onto a glass substrate with anisotropic conductive film (ACF). With the higher demand on the display capacity, the stability of COG interconnection under the finer pitch package becomes more important. Currently bump conductive particle counting is used to assess the interconnection quality. However, it cannot verify the interconnection conductivity promptly and would be labor intensive in case that all the bumps were required to inspect. In this paper, a conductivity inspection system based on optical microscope is presented. The resistance prediction model used in the system is based on particle conductive area and has been verified with the published experimental data, which revealed it could calculate the resistance value with the mean of particle conductive area and the number of conductive particles trapped by bump. The model is more accurate than the existed models. A visual inspection system for bump interconnection is developed in this study to compute the particle conductive area and the particle number. The system has been tested with 228 bump images of the ACF-COG modules, and the predicted resistance, 163 m X, is in the common range. The results show that the conductive area (CA) can be an integral index of bump conductivity and comply with a normal distribution. The resistance model and inspection system provide guidance for the COG interconnection conductivity online testing and will benefit the COG assembly reliability. 2012 Elsevier Ltd. All rights reserved.

Evaluation of a double-layer anisotropic conductive film (ACF) for fine pitch chip-on-glass (COG) interconnection

Anisotropic conductive films (ACFs) consist of conductive particles and adhesive resins and have been extensively used for chip-on-glass (COG) interconnection in FPDs (flat panel displays) such as LCD (liquid crystal display) for the last decades. It is one of the highest density packaging processes at present. However, the need for higher resolution and larger capacity displays is driving the need for finer pitch interconnection. In order to meet these requirements, many approaches were developed by researchers. Since it is difficult to capture enough particles on the small bumps while securing the insulation between the adjoining terminals, currently two types of ACF are widely adopted: an ACF containing conductive particles coated with insulated layer; a double-layer ACF that consists of an ACF layer and a NCF (nonisotropic conductive film) layer. However, both of them have the inevitable limitation. An ACF which has a double-layer structure and the ACF layer consisting of conductive particles coating with insulated layer was investigated in this paper. We carried out the COG experiments with test TC and glass under several different contact areas, another kind of normal single-layer ACF was also adopted which has the higher particle density than the novel ACF for comparing. It was found that the number of particles trapped between the bump and the corresponding pad increased obviously. A thermal cycling test and a high temperature humidity test were also taken, the ACF showed more acceptable contact resistance, and no short was found.

Thermal and Mechanical Loading Effects on the Reliability of COG-ACF with Thin Glass by FEA

The chip on glass (COG) technique has been developed for excellent resolution and high quality liquid crystal display (LCD) panels for several years. In telecommunication field, the glass of LCD with 0.3mm thickness is used in the real product soon. However, many serious manufacturability and reliability issues were observed in the previous studies. In those, warpage of LCD cell is one of the common concerns, for warpage of IC and glass would induce delamination between the interface of gold bump/anisotropic conductive film (ACF) and ACF/glass. The critical parameters of COG process must be closely investigated to achieve minimum warpage. In this paper, based on the finite element analysis (FEA) tool, ABAQUS, we established numerical model of LCD cell, which was verified by warpage measurement system, by taking the detailed interconnect structure into consideration and the equivalent particle method to simulate ACF bonding process and investigate IC warpage under thermal-mechanical loading. We studied heat transfer, critical stress inducing delamination and equivalent particle deformation in the global and local scale. We also explored the effects of bonding force, bonding head temperature and stage temperature on warpage. From the simulation results, warpage are primarily determined by the difference of bonding head temperature and stage temperature

A Registration Method for 3D Point Clouds with Convolutional Neural Network

Viewpoint independent 3D object pose estimation is one of the most fundamental step of position based vision servo, autopilot, medical scans process, reverse engineering and many other fields. In this paper, we presents a new method to estimate 3D pose using the convolutional neural network (CNN), which can apply to the 3D point cloud arrays. An interest point detector was proposed and interest points were computed in both source and target point clouds by region growing cluster method during offline training of CNN. Rather than matching the correspondences by rejecting and filtering iteratively, a CNN classification model is designed to match a certain subset of correspondences. And a 3D shape representation of interest points was projected onto an input feature map which is amenable to CNN. After aligning point clouds according to the prediction made by CNN, iterative closest point (ICP) algorithm is used for fine alignment. Finally, experiments were conducted to show the proposed method was effective and robust to noise and point cloud partial missing.

A Novel Conductive Particle Dispersing Method via EHDA for POB-COG Packaging

With the continuous evolving of Liquid Crystal Display (LCD) panels and Integrate Circuits, conventional Anisotropic Conductive Film (ACF) has confronted with two paradoxical requirements: higher conductive particle density, while avoid serious aggregation. To solve this problem, a wafer level pretreatment technique, Particle on Bumps (POB), was proposed to renovate the usual Chip on Glass (COG) technology. Although we have demonstrated its feasibility and reliability, a fundamental route towards fully accomplishing the POB technique still needs to be paved out, that is the dispersing of conductive particles. Herein we introduced the method of Electrohydrodynamic Atomization (EHDA). EHDA is a technique which ultra-fine droplets can be induced due to electrical force. We build up a prototype apparatus for particle dispersing. With the help of high speed camera, we are able to verify certain behaviors and characteristics of different modes during the EHDA process. The deposition patterns generated under different spraying modes are observed and compared; the stable cone jet mode is therefore chosen as the operational state during our following deposition process. We investigate the connections between onset voltage and relevant spray parameters to optimize our EHDA configurations. By implementing the particle dispersing experiment via stable cone jet EHDA, we successfully obtained a pattern with high densely and uniformly distributed particles.

A Deep Learnable Framework for 3D Point Clouds Pose Transformation Regression

Object detection and classification for point clouds with deep learning module has gained some substantial progress recently. But there is still a blank for spatial transformation estimation like rigid pose regression with learnable module for point clouds, comparing with other kinds of 2D feature format like range images or RGB images. In this paper, we present a complete framework for point cloud pose regression with the deep learnable module. The method for feeding unordered 3D point clouds to a feature map like 2D images is using the symmetric function, which is the state of the art for object segmentation and classification. The pose regression idea take advantage of the spatial transformation of feature map with neural networks, converging both classification loss and pose regression loss function. We augmented the synthetic sample data by apply randomly sampled rigid pose transformation and Gaussian noise. The whole idea of this study is intuitive and straightforward and we show that the deep learnable module can reduce the rigid transformation loss and the potential application in many fields.

Pose Estimation with Mismatching Region Detection in Robot Bin Picking.

3D object detection and pose estimation based on 3D sensor have been widely studied for its applications in robotics. In this paper, we propose a new clustering strategy in Point Pair Feature (PPF) based 3D object detection and pose estimation framework to further improve the pose hypothesis result. Our main contribution is using Density Based Spatial Clustering of Applications with Noise (DBSCAN) and Principle Component Analysis (PCA) in PPF method. It was recently shown that point pair feature combined with a voting framework was able to obtain a fast and robust pose estimation result in heavily cluttered scenes with occlusions. However, this method may fail in the mismatching region caused by false features or features with insufficient information. Our experimental results show that the proposed method can detect mismatching region and false pose hypotheses in PPF method, which improves the performance in robot bin picking application.

Fast Hierarchical Template Matching Strategy for Real-Time Pose Estimation of Texture-Less Objects

This paper proposes a fast template matching strategy for real-time pose estimation of texture-less objects in a single camera image. The key novelty is the hierarchical searching strategy through a template pyramid. Firstly, a model database whose templates are stored in a hierarchical pyramid need to be generated offline. The online hierarchical searching through the template pyramid is computed to collect all candidate templates which pass the similarity measure criterion. All of the template candidates and their child neighbor templates are tracked down to the next lower level of pyramid. This hierarchical tracking process is repeated until all template candidates have been tracked down to the lowest level of pyramid. The experimental result shows that the runtime of template matching procedure in the state-of-the-art LINE2D approach [1] after applying our fast hierarchical searching strategy is 44 times faster than the original LINE2D searching method while the database contains 15120 templates.

Quantitative Assessment of Uniformity in Particle Distribution

Assessment of particle distribution has a wide range of applications in scientific researches and industrial processes. The knowledge of uniformity, density and agglomeration of distribution is essential for material, medical and many other fields. In this paper, we propose a set of parameters for evaluating the uniformity, density and agglomeration of particle distribution which is based on image processing. Distribution uniformity is evaluated by the value of coefficient of variationcv based on voronoi diagram, and by the eccentricity and deviation angle of particle distribution. The density of distribution is investigated via numerical density and area density which illustrate the area distribution at certain density levels. The assessment of agglomeration is carried out through the separation of particle distribution and the counting of different agglomeration situations. These three aspects of assessment will be demonstrated to fully describe the particle distribution. Microscopic images taken from chip on glassCOG are used to show the assessment process, and the characteristics of distribution can be obtained clearly and comprehensively.

An Automatic Bump Detecting System for LCD COG Module Using Fuzzy Reasoning

The Chip-on-Glass (COG) package, which assemblies the driver IC onto a glass substrate with Anisotropic Conductive Film (ACF), is mainly used in the Liquid Crystal Display (LCD) module. With the higher demand of display capacity, the stability of COG interconnection under fine pitch package becomes more important and difficult. Interconnection test and bump microscopic inspection are commonly used to confirm the viability of the interconnections, and are accomplished manually by engineers, which are repetitious and can merely give out the conductive particle numbers in rough. According to the existent ACF-COG interconnection resistance models, the conductivity of bump and pad can be quantitatively analyzed with the conductive particle contact area. In order to implement an automatic system for the bump microscopic inspection to count the particle number and achieve the distribution of contact areas, it is necessity to realize the bump detecting and extracting automatically. In this paper, an automatic bump detecting system using fuzzy reasoning is proposed, and can perform the bump image extraction with a high accuracy. With this system, the particle contact areas and conductive particle number can be analyzed automatically.